DB’s good choices

[GillesD]

Dennis,

I have looked at the data for the 48-49 (with one as the bonus number) combination. You are right: one of these numbers often repeat in the next draw. and if you consider 48-49 coming out as regular numbers (22 times), in 55% of the cases, one of the two did it repeat in the next draw. But I do not have time to check the other 1175 2-number combinations to verify if the 48-49 combination is the best (it should be at least near the top) in that matter.

By the way, could provide some information on how you came out with number 32 as a great announcer for 2 double LD (when considering the 7 numbers). I checked all 49 numbers and the worst announcer is number 33 with 35.6%; Number 32 is better with 42.9% but it is still the 8th worst performer in this matter. How could you pass the 41 other numbers, especially the five numbers with over 60% announcing 2 double LD or better. The best number in that matter is 13 with 64.4%. As mentioned already, over 50% of the draws had 2 double LD or better.

And, Dennis, if you bother answering this posting (and I know you will), could you please leave out the name calling, innuendos and other side remarks. Stick to the facts.

 

[Beaker]

Gilles, if you don't mind, can you post the top 10 announcers and the % for double LD's? This could be helpful Thanks Professor

 

[winhunter]

Gilles, if you don't mind, can you post the top 10 announcers and the % for double LD's? This could be helpful Thanks Professor

OK, Im sure I probably asked this somewhere before, but what is an announcer exactly?



Andrew

 

[Beaker]

OK, Im sure I probably asked this somewhere before, but what is an announcer exactly?

Andrew

Andrew check this 5 posts down for a post by Beaker for a brief description.
Basically, announcer = predictor. Numbers can be announcers or be announced by

 

[Dennis Bassboss]

Dennis,

I have looked at the data for the 48-49 (with one as the bonus number) combination. You are right: one of these numbers often repeat in the next draw. and if you consider 48-49 coming out as regular numbers (22 times), in 55% of the cases, one of the two did it repeat in the next draw. But I do not have time to check the other 1175 2-number combinations to verify if the 48-49 combination is the best (it should be at least near the top) in that matter.

By the way, could provide some information on how you came out with number 32 as a great announcer for 2 double LD (when considering the 7 numbers). I checked all 49 numbers and the worst announcer is number 33 with 35.6%; Number 32 is better with 42.9% but it is still the 8th worst performer in this matter. How could you pass the 41 other numbers, especially the five numbers with over 60% announcing 2 double LD or better. The best number in that matter is 13 with 64.4%. As mentioned already, over 50% of the draws had 2 double LD or better.

And, Dennis, if you bother answering this posting (and I know you will), could you please leave out the name calling, innuendos and other side remarks. Stick to the facts.

Thanks Gilles! And you are right I will only sticks to the facts my apology for my babling tendancy it is I guess part of my nature and please accept my excuses for that...but I will listen to you and I'll stick to the facts!Thanks for all these infos! I'm still looking at many of these numbers to see if some are announcing some specific LDs..When done with it I will post the results of my search! I only came out with 32 because lately I have obtained some success in the past predicting triple LDs 00-01-02-03 it just came out of my head like that because I remembered of it but I might and perhaps should have chosen some others too!

 

[winhunter]

Thanks

That thread explains it!

If I understand correctly, an announcer is nothing more than a trigger. So, within WINHunter you could have a trigger that would eliminate an entire decade based on an annoucer like that "15,2,1".

So what do you guys base annoucer theory on? WINHunter is lacking in the Trigger department.


Andrew

 

[Beaker]

Re: Thanks

That thread explains it!

If I understand correctly, an announcer is nothing more than a trigger. So, within WINHunter you could have a trigger that would eliminate an entire decade based on an annoucer like that "15,2,1".
So what do you guys base annoucer theory on? WINHunter is lacking in the Trigger department.
Andrew

Andrew, there could be literally thousands of these triggers - yet not found. I think this is what we are all looking for

The theory encompasses many things and I don;t think you can look at announcers in isolation - it is more of a step within the strategy to come up with a final set.

I can only speak for my strategy but basically I create a list of numbers to optimize 3 things: repeaters, consecutives, and LD's.
You have a flavour of this in your SW except for LD's.

Note: Add another = Decades

I will look at single digit and pair announcers to "filter" potential candidates both from the entire history and from the last draw.
Example, the last draw was 14-20-22-41-43-49 B16. The question is what did 14 announce overall in the history and what did each pair announce. Then within the history, look at sub-history of 2,3,4,5 numbers from the last draw to see if there is anything interesting.

I must acknowledge Dennis Bassboss here. He was the first to publish the concept of announcers of "history within history".

Announcers might put you in the ballpark but you must look at other things including repeaters, consecutives, and LD's

My .02

 

[Dennis Bassboss]

Re: Re: Thanks

Andrew, there could be literally thousands of these triggers - yet not found. I think this is what we are all looking for

The theory encompasses many things and I don;t think you can look at announcers in isolation - it is more of a step within the strategy to come up with a final set.

I can only speak for my strategy but basically I create a list of numbers to optimize 3 things: repeaters, consecutives, and LD's.
You have a flavour of this in your SW except for LD's.

I will look at single digit and pair announcers to "filter" potential candidates both from the entire history and from the last draw.
Example, the last draw was 14-20-22-41-43-49 B16. The question is what did 14 announce overall in the history and what did each pair announce. Then within the history, look at sub-history of 2,3,4,5 numbers from the last draw to see if there is anything interesting.

I must acknowledge Dennis Bassboss here. He was the first to publish the concept of announcers of "history within history".

Announcers might put you in the ballpark but you must look at other things including repeaters, consecutives, and LD's

My .02

Thanks for the compliment Beaker but as LT already stated earlyer in another post that notion or concept existed way before I spoke about it as for the rest of your post I agree entirely and I strongly believe that the best announcers haven't been found by us yet! This is why I'm always searching and digging for these!And I will continue to search for these even if on my way there I make a few mistakes!

 

[GillesD]

Announcer for 2 double LD or better

Gilles, if you don't mind, can you post the top 10 announcers and the % for double LD's? This could be helpful Thanks Professor

No problem
# % for 2 double LD or better
13 64.4%
01 64.1%
23 63.8%
02 61.3%
27 60.0%
38 59.5%
20 57.5%
39 57.1%
04 56.8%
07 56.8%

and at the othe end
31 36.8%
26 36.1%
33 35.6%

 

[winhunter]

Arg

Ok, So LT wont let me attach files either. Grrr, now I gotta re-type my post.


OK, I see a lottery as MANY different factors, but each is somehow related to the others (rambling). But I saw on another website an interesting method called the ANSELM method. What Dr. Anselm theorized was to predict what direction each draw position took. Graph each position on a piece of paper and you will notice that most draws follow the same rule, that all of the positions take the same direction, they either all increase in value, or they all decrease in value. Now, each position has a MIN/MAX range, but they almost always stay within a given range for that position. Let me explain.

A 6/49 lottery has 6 positions.
Each position is each ball that is drawn, in ascending order.
Position 1 can be 1-44
Position 2 can be 2-45
Position 3 can be 3-46
Position 4 can be 4-47
Position 5 can be 5-48
Position 6 can be 6-49

But, more often than not, these positions stay within a certain Range, thus called Positional Ranges. How much history you use can dictate the average for this range, but the range can never go beyond the above limits.

I think if we can successfully predict direction vectors for positions, we might find that we can eliminate far more numbers. I will give this more thought today, and add some code to the positional range filter for WINHunter.


Andrew

P.S. - From now on, I will [CTRL-C] my post before I click Submit...

 

[Beaker]

Re: Arg

Ok, So LT wont let me attach files either. Grrr, now I gotta re-type my post.
<snip>
I think if we can successfully predict direction vectors for positions, we might find that we can eliminate far more numbers. I will give this more thought today, and add some code to the positional range filter for WINHunter.


Andrew

P.S. - From now on, I will [CTRL-C] my post before I click Submit...

Positional elimination is good but not new - Scott Piel did it in 1996. And if you don't know who he is, he was the developer of Lottoman SW.
I think the biggest "reduction bang for the buck" is in decades. Cracking that nut gives you 9-10 numbers. I'd rather play 6/40 or 6/39 any day of the week. This is where my focus is currently.

The problem with positions is the greyness of the range. Decades are black and white eliminated and you are looking at at least 1 decade missing between 8 and 9/10 draws - call it 8 with the regular numbers. That's high.

 

[winhunter]

Re: Re: Arg

Positional elimination is good but not new - Scott Piel did it in 1996. And if you don't know who he is, he was the developer of Lottoman SW.
I think the biggest "reduction bang for the buck" is in decades. This is where my focus is currently.

So I take it Positional didnt win very often?
I worked for a while on Decades. Trouble was that I never could really settle on a way to easilly and logically divide things up, and what criteria to use to eliminate a decade. My trouble was because Florida went to a 6/53 from a 6/49 (Arg!) So most decades had 9-10 numbers, but the 50's only had 4... So I got stuck and left it alone.


Greyness of the range? With WINHunter, it depends upon how much/little History you use. Which, WINHunter can help figure out how much is just right. I will have to tinker with Positionals and get back to you. Also, it looks like i need to resurrect my decade code and try to port it into a Processor.


Andrew

 

[Beaker]

Re: Re: Re: Arg

So I take it Positional didnt win very often?
I worked for a while on Decades. Trouble was that I never could really settle on a way to easilly and logically divide things up, and what criteria to use to eliminate a decade. My trouble was because Florida went to a 6/53 from a 6/49 (Arg!) So most decades had 9-10 numbers, but the 50's only had 4... So I got stuck and left it alone.
Andrew

The way you use positional is to count the total hits for each number that has appeared in the positions - take the max hit numbers - they are good candidates. For example if 25 has hit in position 2,3,4, and the bonus - pick it. although it may only come in position 3. Scott Piel predicts positional candidates and ranks them. So, 25 might rank 6 in position 2, 5 in position 3 and 2nd in position 4 - similar to your weighting system. He also provided a feature to predict actual positions (very restrictive) or any position.

Well, Andrew, we need to mine the history to find the trigger(s) for that decade reduction

You are right about the other lottos - 6/53 or 6/45 would be a little bit of a pain. 6/49 is great - I'll take it

 

[winhunter]

Re: Re: Re: Re: Arg

You are right about the other lottos - 6/53 or 6/45 would be a little bit of a pain. 6/49 is great - I'll take it

 

[Beaker]

Re: Re: Re: Re: Re: Arg

6/49 - the best

By the way Andrew, keep up the great work - this could lead to something really cool. I enjoy seeing what the SW does. Too bad I don;t speak VB

 

[winhunter]

Re: Re: Re: Re: Arg

The way you use positional is to count the total hits for each number that has appeared in the positions - take the max hit numbers - they are good candidates. For example if 25 has hit in position 2,3,4, and the bonus - pick it. although it may only come in position 3. Scott Piel predicts positional candidates and ranks them. So, 25 might rank 6 in position 2, 5 in position 3 and 2nd in position 4 - similar to your weighting system. He also provided a feature to predict actual positions (very restrictive) or any position.

But there is a slight problem with the above method. 1 & 49 will only appear in 1 position, ever. 2 & 48 in 2 positions, 3 & 47 in 3 positions. 6-44 can appear in all 6 positions, but do they? 6 will only appear in the 6th position if the other 5 numbers are 1,2,3,4,5....


Andrew

 

[winhunter]

Re: Re: Re: Re: Re: Re: Arg

6/49 - the best

By the way Andrew, keep up the great work - this could lead to something really cool. I enjoy seeing what the SW does. Too bad I don;t speak VB

Here is the code for the single hit processor. This simply counts the hits for a given Start/Stop range of drawings...

'This routine simply totals the number of times
'a single ball has been drawn for the number of drawings
'specified
Public Sub Calc_Single_Scans()
Dim mDigit As Long
Dim mSum As Long

'We must have a Drawings Object to continue!
If mDrawings Is Nothing Then Exit Sub
'Now we must have a start and stop!
If mStart = 0 Then
mStart = 1
End If
If mStop = 0 Then
mStop = mDrawings.Count
End If


'Reset Values Before Calculating!
mMinSum = 10000
mMaxSum = 0
mAvgSum = 0

For i = mStart To mStop
mSum = mDrawings.Item(i)
For n = 0 To mDrawings.Drawn - 1
mDigit = mDrawings.Item(i).Numbers
'mSum = mSum + mDigit
mSingleHit(mDigit) = mSingleHit(mDigit) + 1
Next n
mAvgSum = mAvgSum + mSum
If mSum < mMinSum Then mMinSum = mSum
If mSum > mMaxSum Then mMaxSum = mSum
Next i
If mStart < mStop > 0 Then
mAvgSum = mAvgSum / (mStop - mStart)
End If
mLastStop = mStop

End Sub


Not too complicated is it?


Andrew


-It lost the formatting-

 

[Beaker]

Re: Re: Re: Re: Re: Arg

But there is a slight problem with the above method. 1 & 49 will only appear in 1 position, ever. 2 & 48 in 2 positions, 3 & 47 in 3 positions. 6-44 can appear in all 6 positions, but do they? 6 will only appear in the 6th position if the other 5 numbers are 1,2,3,4,5....
Andrew

You have identified both a problem and a benefit and it has to with the 1 and 49 numbers. You are right, 1 can only appear in position 1 (but I include the bonus number in the config which can but rarely does have the number 1.) and 49 can only be in position 6 and because of this you don't see them that often. If I see the 1 in position 1 it usually has a very low rank and I don;t see the 49 that often in position 6. However, if I dosee these numbers, I pay particular attention because they are very specific predictions which can be valuable.

You have to be careful with position 1 - its the key. If you get that wrong, everything else is wrong.

 

[Dennis Bassboss]

Thanks Gilles,Beaker and Winhunter for this most appreciated lotto discussion it is mostly interesting and at the same time very refreshing!

 

[Beaker]

Thanks Gilles,Beaker and Winhunter for this most appreciated lotto discussion it is mostly interesting and at the same time very refreshing!

Thanks Dennis. Now if I can put all that talk into some good picks, I can get out of my slump