Winnalotto-Sangoma 2

Patrick123

Patrick123

Member
Bonjour everyone,

I have started a new thread in the International & State Lotteries section for the SA Lotto predictions. This way we should be able to keep this thread a bit cleaner and relate only to matters regarding WinnaLotto.

Feel free to add your predictions there. Any comments and questions how the numbers were selected would be welcome.

Regards
Patrick
 
B

black prince

Member
Patrick123


Originated from Patrick
I will upload an update for WinnaLotto later on today. BP, I've added in a complete combination generator, it stores everything in memory, I'd suggest not to generate mega sized combinations it should handle 1 million or 3 no problem. I'll add in a feature to load your own file as well.
Click to expand...

Thanks Patrick, this helps me keep all my work mostly in Winnalotto. As you
refine and tweak Winnalotto it will mature and become a very valuable tool for
serious lottery players. :thumb: I have a fairly good computer with 2.8Ghz
intel processor, 4 gig of memory, and 200gig SATA hard drive. It's a Dell.
I can store quite a bit in memory. ;)

-BP
 
Patrick123

Patrick123

Member
Hi BP,
That should handle everything quite comfortably.
I want to now start looking at emulating some of Bright5 & Bright6's features, so that more and more intelligent selections are kept/made :)

Regards
Patrick
 
B

black prince

Member
Patrick123


Originated from Patrick123
Hi BP,
That should handle everything quite comfortably.
I want to now start looking at emulating some of Bright5 & Bright6's features, so that more and more intelligent selections are kept/made

Regards
Patrick
Click to expand...

It appears the 'Wheel Tab' may no longer be necessary. I'm truely impressed
with the current 'Purge Screen', but more intellegent dynamic filters need
investigating. Applying 'Standard Deviation' ranges based on past draws to
filter could remove most of the low probability draws. This reduces the need
for so many 'Static Filters' like odd/even, primes, decades, distance, ending
digits, sequential numbers, on and on. We are working to extract high
probability draws, but sometimes and hopely not too often, a weird draws slips
thru. If the maximum filters produce 80% to 90% then this is our goal, and
not the improbable 100% winner every time.

-BP
 
B

black prince

Member
Patrick123

I use this formula to calculate SDP ranges from past winning draws. This
filters out about 30% to 40% of my draws and is much less distructive in removing
draws than static filters. You run this against your history file and determine
where the most draws come from. Then set the ranges as a filter.

Standard Deviation Population

Draw: 1, 2, 3, 4, 5

1) S1 = [(1)^2=1 + (2)^2=4 + (3)^2=9 + (4)^2=16 + (5)^2=25] = 55

2) S2 = [1 + 2 + 3 + 4 + 5] = 15

3) S3 = [(S2 * S2) / 5)] = [(15 * 15) / 5] = 45

4) S4 = [S1 - S3] = [(55 - 45)] = 10

5) S5 = [SQR(S4)=3.16 / SQR(5)=2.24] = 1.42

This is the lowest SDP

-BP
 
Patrick123

Patrick123

Member
Hi BP,

I agree about the Wheel tab, but wonder if I should leave it in there for people who prefer to use that method?
I was actually thinking of adding in a single click button as an extra option to chop out all the rough stuff. I'd like to leave the rest in as I feel once again some people might prefer the finer tuning for their purges.

Once we get the more intelligent filters working, I'll place the current ones basically on the 'back-page' for anyone who is might be inclined to want to continue to experiment with them.

Regards
Patrick
 
Patrick123

Patrick123

Member
Hi BP,

Thanks, I'll look at that. I want to experiment & see if I can get some method defined to that most of these dynamic filters can be loaded and run similar to the column filters. That would give us all that freedom to experiment again.

In fact I'll most probably use your filter as a prototype to try and work out the 'language' for these filters.

Regards
Patrick
 
B

black prince

Member
Patrick123

Tried the filter 'Very Hot Numbers' and it was right on the money predicting
13 would appear in the next CASH5 lottery. It did :thumb: I'm very
excited about these new filters. Can you recommend a few more.

I like loading filters. I also, have a few favoriate wheels I wish I could load in the Wheel Screen.

Thanks

-BP
 
Patrick123

Patrick123

Member
Hi BP,

I must say there's nothing better than seeing your win average climb steady:bounce:

If the Yo Yo Up High Prob kicks out at trip (usually only follow the trips with this one) it is virtually guaranteed to hit within that draw or the next.

I want to follow up on the ability of loading wheels defined by other people, e.g. Nick Koutras etc. I also want to bring the wheeled numbers onto the Purge Tab.

Regards
Patrick
 
Patrick123

Patrick123

Member
Hi BP,

Thou wish is my command ;). The format is the same as Nick Koutra's so there should be no problem.

I have uploaded the updated version. This essentially adds, the ability to load and evaluate lottery wheels. I must still incorporate this into the purge screen... watch this space :)... , another filter was added to allow you to define a minimum value for a column on sub-tab purge 2, with the Totals filter moved to sub-tab purge 3.

You'll notice a purge 4 tab. This is underdevelopment and does not work yet! This is how I envisage the initial auto-purging of most of the rows generated. The sliders are based on the bell curve and pre-set at 1 std deviation equal to 68%, you'll thus have the capability of skewing the curve to whatever you gut feel thinks is going to be best. This is mainly there to get a bit of feed-back from everyone.

The wheeling tab at this stage has a preset bias for calculating the best wheel. i.e. the Match 6 has a higher preference to match 3. I envisage using the bias tab to allow you to define what is more important when generating your wheel.

Regards
Patrick
 
B

barge

Member
Greetings All

Having been discovering Portugal for the past few months, I have just discovered the revitalised (?) site, and am trying to catch up with the revised prog! It is good to see you all back in harness........if that is the right word!!
 
B

black prince

Member
Patrick123


Originated from Patrick123
Hi BP,

Thou wish is my command . The format is the same as Nick Koutra's so there should be no problem.

I have uploaded the updated version. This essentially adds, the ability to load and evaluate lottery wheels. I must still incorporate this into the purge screen... watch this space ... , another filter was added to allow you to define a minimum value for a column on sub-tab purge 2, with the Totals filter moved to sub-tab purge 3.

You'll notice a purge 4 tab. This is underdevelopment and does not work yet! This is how I envisage the initial auto-purging of most of the rows generated. The sliders are based on the bell curve and pre-set at 1 std deviation equal to 68%, you'll thus have the capability of skewing the curve to whatever you gut feel thinks is going to be best. This is mainly there to get a bit of feed-back from everyone.

The wheeling tab at this stage has a preset bias for calculating the best wheel. i.e. the Match 6 has a higher preference to match 3. I envisage using the bias tab to allow you to define what is more important when generating your wheel.

Regards
Patrick
Click to expand...

I have a lot of back testing to do and I'm glad that most, if not all, can be
done in the new Winnalotto. You have given us lots of options, which is
just what we needed. ;)

-BP
 
Patrick123

Patrick123

Member
Hi BP,

I see these next few weeks are going to be a bit of a roller-coaster ride, hold onto your hats :)

Hi barge,
I also only got back from my projects a short while back. At this stage WinnaLotto is start to perform above expectations, and there is still quite a whack to do. We're now at the level of starting to define the tertiary filters.

Regards
Patrick
 
B

black prince

Member
Patrick123 and Springbok

Today's Winners (5/17/08):

--------------------------------------------------------------------

Treasure Hunt (5/30):

Winnalotto's Picks -

1) Very Hot Numbers --> 21
2) Fullhouse6 --> 6:13:15, 7:13:30
3) Springbok2 --> 7, 15, 25, 30
4) My_Columns --> 1:18, 5:19, 1:17
................. 15:24:27, 14:25:30, 4:22:24

Winning Draw: 5 - 14 - 19 - 27 - 28

---------------------------------------------------------------------

Cash5 (5/43):

Winnalotto's Picks -

1) Fullhouse6 --> 24:28, 28:35
2) YoYo Up -----> 10:18, 40:43 ... 11:18:36, 25:29:42
3) My_Columns --> 9:15, 1:11, 2:20
................. 9:15:29, 19:29:35, 4:9:19

Winning Draw: 6 - 14 - 34 - 35 - 42

---------------------------------------------------------------------

PowerBall (5/55 +PB 1/42):

Winnalotto's Picks -

1) YoYo Up High --> 6:50:54, 14:15:17, 15:17:54, 17:27:34
2) Springbok2 ----> 12, 14, 15, 16, 47
3) Fullhouse6 ----> 27:40
4) My_Columns ----> 45:52, 8:23, 13:32
................... 45:46:52, 16:31:40, 25:45:46

Winning Draw: 12 - 14 - 15 - 21 - 43

--------------------------------------------------------------------

What's interesting about these results is that Treasure Hunt have been returning
4 out of 5, and 5 out 5, fairly regularly. Even when back testing CASH5 and
PowerBall only return 3 out of 5 or less. This leads me to believe the column
filters need to be customized to the game's winning patterns. I'm just lucky
that it works so well with Treasure Hunt. My_Columns are 'N+x>=H', 'MaxVal',
and 'Double Hit Perc'. How can I tweak these column filters for CASH5,
PowerBall or any game?

-BP
 
B

black prince

Member
Patrick123


.
.
.
MaxVal
False
MaxVal
30
This is the Highest that any skip cycle has gone for the Number
True
False
60
0
0
4
.
.
.
N+x>=H
False
N+x>=H
50

True
False
60
0
0
6
.
.
.
Double Hit Perc
False
Double Hit Perc
120

True
False
60
0
0
13
.
.
.
Click to expand...
The above is the code from the Base filter highlighting 'N+x>=H', 'MaxVal', and 'Double Hit Perc'.
How can I tweak this column filter for CASH5, PowerBall, and other lottos?

-BP
 
Patrick123

Patrick123

Member
Hi BP,

The best way to do this is to evaluate the skip cycles as well as those fields. You will not be able to modify them as they are pre-set within the program, but you are able to create a new filter.
For example, on the columns tab, load Yo-Yo Up High Prob. and modify Column 30 to read:
#20 > #1
and
#21 < #1
and
#22 > #1
and
#23 < #1
and
#24 > #1
and
#51 < #1
and
#52 > #1
and
#53 < #1
and
#8 > 0
and
#5 >= 50
Click to expand...

Save this filter as something else, load up about 7 historical draws, load this filter, then click evaluate and see how it performs. You can then modify the 50 to a higher or lower figure until it starts producing some results.

If you produce too many results, sort by the In Draw column and try and see if there is anything in common with the winning balls. You then add this criteria in by appending it to column 30 or adding a column 31 as a filter type with this logic.

Regards
Patrick
 
B

black prince

Member
Patrick123

Since filers are relying heavily on medians and Skips, I compared Winnalotto's
median's against BRIGHT's medians. The skips are perfect, but some medians
are correct and some are off. (W) is for Winnalotto and (B) is BRIGHT5.

Winnalotto's Medians vs BRIGHT5 Median's

Treasure Hunt (5/30). 377 history draws up to 5/17/08:
This is from FREQ5 report in BRIGHT5 on the same draws used in Winnalotto.

# ......W.....B

01 -> (4) / (3)
02 -> (4) / (3)
03 -> (6) / (7)
04 -> (2) / (2)
05 -> (4) / (3)
06 -> (4) / (3)
07 -> (4) / (4)
08 -> (3) / (5)
09 -> (3) / (3)
10 -> (3) / (5)
11 -> (4) / (5)
12 -> (4) / (4)
13 -> (4) / (2)
14 -> (4) / (3)
15 -> (4) / (3)
16 -> (3) / (2)
17 -> (4) / (5)
18 -> (4) / (4)
19 -> (3) / (4)
20 -> (5) / (5)
21 -> (2) / (1)
22 -> (3) / (3)
23 -> (2) / (2)
24 -> (2) / (2)
25 -> (3) / (2)
26 -> (4) / (3)
27 -> (3) / (3)
28 -> (4) / (5)
29 -> (2) / (3)
30 -> (3) / (2)

I used the Balls button and Base filter to get these medians for comparisons.
How are medians calculated in Winnalotto? Are you using average or mid-point?

-BP
 
B

black prince

Member
Patrick123

Opps!! The default range of Winnalotto is 200. I adjusted this to 1,000
to load all history and the result is as follows:

Winnalotto's Medians vs BRIGHT5 Median's

Treasure Hunt (5/30)

# ......W.....B

01 -> (4) / (3)
02 -> (4) / (3)
03 -> (6) / (6)
04 -> (2) / (2)
05 -> (4) / (4)
06 -> (4) / (4)
07 -> (4) / (4)
08 -> (3) / (3)
09 -> (3) / (3)
10 -> (3) / (3)
11 -> (4) / (4)
12 -> (4) / (3)
13 -> (4) / (4)
14 -> (4) / (4)
15 -> (4) / (3)
16 -> (3) / (3)
17 -> (4) / (4)
18 -> (4) / (3)
19 -> (3) / (3)
20 -> (5) / (5)
21 -> (2) / (2)
22 -> (3) / (3)
23 -> (2) / (2)
24 -> (2) / (2)
25 -> (3) / (3)
26 -> (4) / (3)
27 -> (3) / (3)
28 -> (4) / (4)
29 -> (2) / (2)
30 -> (3) / (2)

Sorry about that. BRIGHT5 calculates the median, not the average, so the
differences are small. In forecasting could this make a difference?

-BP
 
You must log in or register to reply here.

Sidebar

Top