Stats/WNH Segs

[Red Devil]

Stan,

I don't know if this is supposed to happen but when using Stats/WNH Segments lets just say you select for column-1 Sement 321, select filter and preview what is left in the package, when you cancel out of this screen and go back into Segments Table screen, select another column, the occurences are now incorrect. To refresh them you have to press the "Calculate" stats button again?

Red Devil

[Red Devil]

Stan

I think that some of the stats for the segments table are a little misleading. When you consider that the segments are based on the last 11 historic occurrences of any ball appearing you have to consider that for the majority of balls to be drawn at least 11 times probably took over 140 draws. These first draws of the lottery should then be discarded to get realistic stats. For example if you look at column 10 in the Stats/WNH Segments Table it shows that there have been 80 occurrences of Segment 6-0-0 however, 61 of these occurrences were in the first 61 draws of the Lotto history and the average skip 15 does not reflect a true value for this segment, a more realistic skip would be 61, (number of draws/occurrences since draw history had settled). See attached, what do you think?

Red Devil

[stan]

[quote=Red Devil:1172254581]
Stan,

I don't know if this is supposed to happen but when using Stats/WNH Segments lets just say you select for column-1 Sement 321, select filter and preview what is left in the package, when you cancel out of this screen and go back into Segments Table screen, select another column, the occurences are now incorrect. To refresh them you have to press the "Calculate" stats button again?

Red Devil
[/quote]

that's a bug (the stats will include occurrences found by the filter...), will fix it in the next build.

[stan]

[quote=Red Devil]
Stan

I think that some of the stats for the segments table are a little misleading. When you consider that the segments are based on the last 11 historic occurrences of any ball appearing you have to consider that for the majority of balls to be drawn at least 11 times probably took over 140 draws. These first draws of the lottery should then be discarded to get realistic stats. For example if you look at column 10 in the Stats/WNH Segments Table it shows that there have been 80 occurrences of Segment 6-0-0 however, 61 of these occurrences were in the first 61 draws of the Lotto history and the average skip 15 does not reflect a true value for this segment, a more realistic skip would be 61, (number of draws/occurrences since draw history had settled). See attached, what do you think?

Red Devil
[/quote]

that's a common problem for all wn history stats. it takes a few draws before the stats even out.
try using 'latest 1000' draws in your wn selection instead of the 'all draws', the stats should be more accurate then.

[Red Devil]

[quote=stan]
[quote=Red Devil:1172254581]
Stan,

I don't know if this is supposed to happen but when using Stats/WNH Segments lets just say you select for column-1 Sement 321, select filter and preview what is left in the package, when you cancel out of this screen and go back into Segments Table screen, select another column, the occurences are now incorrect. To refresh them you have to press the "Calculate" stats button again?

Red Devil
[/quote]

that's a bug (the stats will include occurrences found by the filter...), will fix it in the next build.
[/quote]

Stan,

That's what I thought, OK for next build

[BrianPA]

Stan
In build 929, The stats page WNHsegments there is a problem. In col. -2 segment 4-0-1 is showing me 0 occurrences when this is the correct segment for the winning ticket, and the winning ticket is still in the package. Also when I filter from the stats page col. -8 the segment 3-2-0 which is also the winning segment it filters out the winning ticket. I tried updating and recalculating the stats to see if it was something I was doing wrong, but these two problems are still there.....Sorry!

After further review......There are more issues with the filtering from the stats page/WNH segments. Just tried removing a segment that was not the winning segment and it removed the winning ticket.

[BrianPA]

Here lies the problem. :finger: I have found that the summary table (segments selected) in the WNHistory page is showing different segments than if I load winning numbers in the package and going to stats page and selecting winning numbers stats. :wand:

I have the default segment size set at 13/26 because I'm playing a 5/39 lottery. Not sure if that has anything to do with it?

[sysp34]

hi BrianPA

maybe this link thread helps

http://www.expertlotto.com/en/forum/forum.php?req=thread&id=481&page=4&nocount=1

good luck

[BrianPA]

Well.....now I'm really confused. Was just doing some backtesting and the segments were way off. What do I use as the winning segment for each column when doing backtesting? Maybe just load the winning ticket in the package and do a stat page segments calculation??

[george]

[quote=BrianPA]
Stan
In build 929, The stats page WNHsegments there is a problem. In col. -2 segment 4-0-1 is showing me 0 occurrences when this is the correct segment for the winning ticket, and the winning ticket is still in the package. Also when I filter from the stats page col. -8 the segment 3-2-0 which is also the winning segment it filters out the winning ticket. I tried updating and recalculating the stats to see if it was something I was doing wrong, but these two problems are still there.....Sorry!

After further review......There are more issues with the filtering from the stats page/WNH segments. Just tried removing a segment that was not the winning segment and it removed the winning ticket.
[/quote]

i saw the same problem when i tested Maryland strategy, i pick one draw for test and i filter the package with the same segment like the draw had but i loose the jackpot in this way

(from -2 or -3 i loose the jackpot)

example

for this prediction http://www.expertlotto.com/en/forum/for ... ead&id=545 you have to filter like:
0 500
-1 410
-2 410
-3 410


but if i check now in wnh segment for 2.3.2007 is
0 410
-1 410
-2 410
-3 410


the problem is if you filter with 0 410 you loose the jackpot :wand:

[PadawanLotto]

george, I found no problem as you described with build 929.
Make sure you update everything and that you draw dates are set to the ones that you are working with. One more thought, when looking at the segment results the segment balance must be the same as the balance that you are using for filtering.

Segment results in the Summary Table can be confusing.

[george]

i checked my results and the draw from 3 was inaccurate, i will check again

ok i checked again in stats for 0 i have 410, if i filter package with 0 410 i loose the jackpot

[stan]

[quote=PadawanLotto]
Segment results in the Summary Table can be confusing.
[/quote]

i'm aware of this but don't know a better way to represent that data - any suggestions anybody?

[PadawanLotto]

[quote=stan]
[quote=PadawanLotto]
Segment results in the Summary Table can be confusing.
[/quote]

i'm aware of this but don't know a better way to represent that data - any suggestions anybody?
[/quote]

No, actually that is the way it should be. I have been keeping track of simulation tables in a spreadsheet and changing one of the columns to see the actual difference of the next drawing for comparence. The actual difference appears in the simulation table that is the drawing before.

[george]

what i made now was to backtest the draw from 2.3 with the segments from 1.3 and the result was surprising for me ... i didn't loose the jackpot