Still looking to put together macro for sum of skips!

[mirage]

Quote originally posted by Snides

about the inner/outer and rows/columns filter.. you are in ontario, so look at your play slips.. See how it's divided into a 7x7 grid? the outside of that grid is your "outer", the inside is the "inner", and rows/columns, that should be obvious now, you mentioned you were more "visual" than mathmatical before.. each row and column would be a filter. Generally you get 2 or less per row or column, with a few occassions of 3 in some (including bonus).

Ok, I understand. Someone else brought this subject up in previous post on this thread, and it wasn't addressed. My personal feeling (imo) is that it is not a particularly useful filter, especially in the grid or row format. Amounts to basically the same filter as "small numbers vs big numbers", and is indirectly also a "spread" filter. However, my opinion only, if you think its worthwhile, it's your program.

I have a number of ideas about filters that through observation(imo), I think are useful. I will address these in posts to come.

Btw, your programming skills must be considerable, if you can manage sum of skips filter to specifications, (You have been able to manage sum of skips without duplicates?, and that isn't a lot of computing time?), plus fairly easily any other filter that one might desire!!

If you can pull this off in a user friendly format, my hat's off to you, and my admiration is immense!

However, not to overlook the rest. There appears to be a number of you on this board who are also talented and skilled and I hold you all also in great respect and esteem!

 

[mirage]

Quote originally by Snides

One question i would like to ask first, what is your preferred screen resolution? I'm trying to design it for 800x600, but if you use 1024x768, that would be great, i can increase the size of the form and add more stuff in then..

I have 1024X768 screen resolution. However, if the font size is too tiny, it becomes a bit of a problem. Thanks!

 

[mirage]

Originally posted by Peter
Well lets plant the seed, and spill the beans, and lets see what grows.

Whew!!! One problem at a time here! I'm still addressing filters and if I'm if I'm focusing there first I need to complete the job. I also have a life that needs attending to, so I don't know how much I will be able to even get to today! I haven't had a chance to check out my idea at all! Who knows, maybe it's been thought of before and is really corny/and or impractical! I don't want to be laughed off the board. I know - you are challenging my mettle and credibility. Well, patience, if you don't mind!
Stay tuned. I'm sure you will!

 

[mirage]

Originally posted by Snides on 17/06/2004
As an example, I requested a sum of skips of 50 for draw number 1852 in 6/49. It listed all 294322 different combinations that totalled that sum of skips, and of course the jackpot was in the list. 9 10 22 38 43 and 44 were all there.. I don't think it captured any 5+B wins though.. Sum skip totals from 24 to 58 were all above 200,000 combinations.

This is not completely clear to me. Are you saying that the spread between sum skip totals of 24 up to 58 were all above 200,000 combinations each?
I guess one could infer that those sum skips with greatest number of combos might have slightly greater tendency to include the winning numbers, maybe?

 

[mirage]

possible filters to use

Originally posted by Snides

...about the inner/outer and rows/columns filter.. you are in ontario, so look at your play slips.. See how it's divided into a 7x7 grid? the outside of that grid is your "outer", the inside is the "inner", and rows/columns, that should be obvious now, you mentioned you were more "visual" than mathmatical before.. each row and column would be a filter. Generally you get 2 or less per row or column, with a few occassions of 3 in some (including bonus).

Snides,

(My lack of sleep is starting to catch up with me...).
Re: your suggestion as a filter noted above. Frankly, I'd never thought about rows/columns before, thought it was just a pattern avoiding technique in an effort to randomize the plays. (seems PAB was looking for macro for this in "Maggie" thread.) A cursory look at historical data for Ont 6/49 seems to support this as being a good filter. Therefore,
1) rows and columns -yes!
2) However, another useful filter is combos with at least 2 or more consecutive numbers - I think this may run contrary to filter #1), at least sometimes. Consecutives seem to show in 50% to 60% approx of winning draws?
3)At least one repeat from previous draw. Happens a bit more than 50% of time?
4)At least one number drawn being directly diagonal to a number drawn in previous draw, elsewhere known as a "leap frogger".
Happens approx 50% to 60% of draws?
5)I went through approx 5 mos worth of draws last night and thought I noticed a pattern of more frequently at least one decade being excluded, rather than there being an even spread. However, checking On 6/49 today, does not seem to be especially significant - maybe 50% of time. (This maybe needs further checking). If you code for consecutives, my inclination would be to exclude at least one decade.
6)At least >than 2 from every decade (as 6 out of 49) but no more than 3.
7)At least 2 last digits but not more than 2 last digits, same - this of course obvious as 6 numbers last digits out of possible 10.
8)At least 4/2 even vs odd, or 4/2 odd vs even, or 3/3 of both.
9)As for inner/outer, this does not seem to be more than 50%, from what I could tell in my small sampling.

Other people might have other findings/opinions.

Some of these filters, i.e. the 50% ones, I think are really just device to eliminate multitudes of combos. Though choices of one vs the other would seem somewhat arbitrary, would still work to good effect for a percentage of draws. (Not that I know what I'm talking about).

How are we doing so far? Would this cut down the numbers significantly? What do you think of above suggestions?

 

[Snides]

This is not completely clear to me. Are you saying that the spread between sum skip totals of 24 up to 58 were all above 200,000 combinations each?
I guess one could infer that those sum skips with greatest number of combos might have slightly greater tendency to include the winning numbers, maybe?

yes, all the sum skip totals from and including 24 to 58 all had more than 200,000 combinations. So slightly more than half the possible 13.9+ million combinations fit into those sum skip totals..

I think;
1) rows and columns are good, I use 0 to 2 per row/column most of the time.
2) consecutives, yes, i use 0 or 1 set of consecutives as a standard filter, i usually do not allow 2 sets (11 12 and 33 34 for example) would not be allowed in my standard filtering.. I also filter out more than 2, so 11 12 13 would not be allowed..
3) one repeat, yes, i use that quite often, though sometimes I will filter out all numbers from last draw if I feel one will not hit..
4) diagonal? leap frogger? not sure what you mean by this one..
5) exclude one decade? yes, I usually use 3 or 4 decades showing up, same as 1 or 2 being excluded. Sometimes I adjust that to making 4 decades manditory, so that all combos have exactly 4 decades.
6) decades, sure, for that i usually use 0 to 2, works in combination with above filter.
7) 2 last digits, ya, that's a good one, many draws have pairs of LD's too.. like 11 21 33 43 as an example.. some draws have 3 last digits the same.. but catching at least 2 of these will help, and reduce the combo's a fair bit..
8) 2/4 3/3 4/2 odd/even, yeppers, I use this as a standard filter..
9) inner/outer, I havn't actually added this one into my wheeler yet, but my plan is to set it similare to the odd/even, with a 2/4 3/3 4/2 filtering.. should help to cut down on the rows/columns filter..

I think all of those suggestions are great ideas, and they should help to cut down on the combinations by a great deal.. I will add them in as i get time, which might be before I get to bed tonight.. or certainly before next week is over...

you also mentioned that this was "my program", well, yes it is, but you're my client, and I'm trying to make it do what you want it to.. if the above is what you want, then i'll make it happen.. I'll also allow you to turn all filters on or off, and allow you to adjust them if you like..

 

[Snides]

Ok, I ran a test looking for a sum skip total of 49 for the last draw with no filters except for less than or equal to 2 numbers with same skip, came up with a little over 277,000 combinations, while still winning the jackpot.

Then I added the following filters;

- Less than or equal to 2 numbers from any decade (0 1 or 2)
- Only 4 decades allowed (no tickets with 1 2 3 or all 5 decades)
- must have one repeater (one number or more, from the last draw
- must have one consecutive (every ticket had to have one set of consecutives, no more and no less)

The results for sum skip total of 49 for the last draw became 19533 combinations. Less than 10% of the original, but the jackpot is gone. Without dumping these tickets to a file for checking wins, or adding in a win checker gizmo to the program, I can't say how well it did, but it looks like a multitude of 3 number wins, and a few 4 number wins scattered about..

Havn't added the other filters in yet.. I will get around to it, but at least you have an update to know how much it can be chopped down..

 

[Snides]

One thing to mention about all these filters.. when you add them all together they can be either good, or bad.. For example, using those filters, and just glancing back a few draws, the first time i notice them working together is draw # 2098 on feb 28 2004. If you had guessed the skip sum total to be 65 correctly, you would had a wheel of 9485 combinations that would have won the jackpot.. not a bad return for less than 10 grand, but you'd hafta get the skip sum total correct..

when the filters do not work together, you should still get multiple smaller wins though.. so some of these filters already added in may need to be broaden or weakened a little.. then with the other filters added, the combinations will still go down.

 

[mirage]

One thing to mention about all these filters.. when you add them all together they can be either good, or bad.. For example, using those filters, and just glancing back a few draws, the first time i notice them working together is draw # 2098 on feb 28 2004. If you had guessed the skip sum total to be 65 correctly, you would had a wheel of 9485 combinations that would have won the jackpot.. not a bad return for less than 10 grand, but you'd hafta get the skip sum total correct..

when the filters do not work together, you should still get multiple smaller wins though.. so some of these filters already added in may need to be broaden or weakened a little.. then with the other filters added, the combinations will still go down.

Good Morning Snides,
Yes, its true - add in all the filters and you cut down the possible combinations significantly, but you have much greater chance of missing the jackpot.

For whatever reason...(a hunch?) I think it's possible most of the filters working together might work a bit better with the Ont 6/49. Just a guess - I have not checked sum skips for that lottery though.

Maybe I'm somewhat out but my sampling of draws for 2 years for the Can 6/49 showed that approx. 80%, or more, of the winning jackpot numbers were within the 24 to 58 sum skip range. Playing combos within that range would cut down number of combos from more than 13,9 mil to a bit more than half that number, as you say.

But running program for ALL the sum of skips within that range plus the extra filters.... impractical?

Other comments:
1) rows and columns are good, I use 0 to 2 per row/column most of the time.
- This filter would probably cut down the combos a lot. Thanks for letting me know about this one.

3) one repeat, yes, i use that quite often, though sometimes I will filter out all numbers from last draw if I feel one will not hit..
- Could this be user's discretion?

4) diagonal? leap frogger? not sure what you mean by this one.
- Here is an example: Last Wed. draw included numbers 46 and 47. Last night's draw (Saturday) included number 45 - diagonal, (or one up or one down), to Wed's 46 on a grid, and number 48 - diagonal, (or one up or one down), to 47 on a grid.

 

[mirage]

Snides,

Oh, one more filter that I use that I forgot to mention (don't know if anyone else uses this one) is filtering for combos that have at least 3 numbers that hit within the last 5 draws (skips 5 or less), inclusive. That is, any with a least 3 or > that hit in last 5 draws.
However, sometimes I vary this...
-----------------------------------------------------------------------------------

After arriving at the number of combos for last Saturday's draw, (177,100, I think it was)the filtering that you did subsequently reduced the possible hits for one sum by a factor of less than 10, i.e. reduced to about 7% of original combos. However it also missed the jackpot. And this was for one skip sum.

One silly strategy would be choose a skip sum that seems to be hitting more frequently, i.e. I think 41 or 35 for last year, were 2 such examples (and my sampling was just over one year, not 2 years as I previously stated), and stick with just that one, in the hopes that it would show again.

My Super-Duper-not-so-brilliant-idea of last Saturday night, in case anyone is still wondering:
Quote "I've just thought of a strategy that I think just might pull off the .....
and I'm hesitant about giving away any more information......." was completely delusional.. as would still result in way too many combos to be practical.

 

[mirage]

Correction to previous quote
After arriving at the number of combos for last Saturday's draw, (177,100, I think it was)the filtering that you did subsequently reduced the possible hits for one sum by a factor of less than 10, i.e. reduced to about 7% of original combos. However it also missed the jackpot. And this was for one skip sum.

- I mean reduced approx same or morer than a factor of ten, sorry about that.(Is that correct?)

 

[Snides]

I think the filter that was cutting out the most winning combinations is the decade one.. I've adjusted it to allow 3 or 4 decades per draw with up to 3 numbers from one decade. By doing this the amount of combinations for the last draw with a skip sum of 49, increased the amount of combinations to 46809. Once the other filters are added, this should come down by a fair bit..

that diagonal/leap frogger, I'll run a test to see how well it performs and add it in.. I want to know if it should be just one leap frog per draw, or up to 2 or 3, etc..

another filter that could be added is position, and position change, but i'll hafta figure out what the best settings for that one would be...

Over the whole history of the 6/49 there were 1644 skip sums within the 24 to 58 range, and 2130 in total. So, 1644/2130 is 77.183098154 % of the draws.. if you go down to skip sum of 20 you boost that to 82.441355%

I'll add in a spot to show you the sum skips for the last 104 draws, so you'll be able to see what was the best for the last year. and i could also show the average for each sum skip for 104 draws. So, say the total history shows a skip sum total of 68 times that it was 41, 68/2130*104 is 3.32 times per year. Then you can look at the last years worth and see that it only happened2 or 3 times so far and you can say "oh look, it's time to play it" and hopefully it works

as for your super-duper delusional idea, maybe it's not as bad as you think.. why not let me know what it is and I can test it to see how valid or invalid it is?

 

[mirage]

I think the filter that was cutting out the most winning combinations is the decade one.. I've adjusted it to allow 3 or 4 decades per draw with up to 3 numbers from one decade. By doing this the amount of combinations for the last draw with a skip sum of 49, increased the amount of combinations to 46809. Once the other filters are added, this should come down by a fair bit..

that diagonal/leap frogger, I'll run a test to see how well it performs and add it in.. I want to know if it should be just one leap frog per draw, or up to 2 or 3, etc..

another filter that could be added is position, and position change, but i'll hafta figure out what the best settings for that one would be...

Over the whole history of the 6/49 there were 1644 skip sums within the 24 to 58 range, and 2130 in total. So, 1644/2130 is 77.183098154 % of the draws.. if you go down to skip sum of 20 you boost that to 82.441355%

I'll add in a spot to show you the sum skips for the last 104 draws, so you'll be able to see what was the best for the last year. and i could also show the average for each sum skip for 104 draws. So, say the total history shows a skip sum total of 68 times that it was 41, 68/2130*104 is 3.32 times per year. Then you can look at the last years worth and see that it only happened2 or 3 times so far and you can say "oh look, it's time to play it" and hopefully it works

Hello Snides,

Rereading a few of my posts for Saturday/Sunday I read/sound like I musta been on good drugs... ! (But only the kind that are legal if anyone cares).

Snides, I think this work and your ideas are great!

- Diagonals/leapfroggers: I think the average is 1.something per draw. If repeats run at about 50% (I don't know exact figure), then diagonals/leapfroggers should be twice that.

- What about the suggestion of filter: include all combos that have skips since last time drawn of 5 or less? I think the average of this should be about 3 numbers - that were last drawn within and including the last 5 draws. Of course includes repeaters and those numbers that were maybe drawn 2 - 4 times in last 5 draws ("hot" numbers) as well as any others.

- You mentioned "position" and "position change" as a filter. What is this?

 

[mirage]

(Originally posted by Snides)

"...as for your super-duper delusional idea, maybe it's not as bad as you think.. why not let me know what it is and I can test it to see how valid or invalid it is?"

Ok, I'll come clean...
It could perhaps be a toy or tool for the already rich or those with resources, I wouldn't recommend it for the average lotto player - you could blow thousands of $$$ trying to get it to work.
Here it is:

1) Instead of all combos for all 49 numbers, why not run a software that creates a file for all combos of 25 numbers (out of the 49) of your own choosing? This would cut down the number of possible combos drastically, to I think, 177,100 in total (is that correct?).

Of course, your chances of having all 6 winning numbers is significantly decreased, but still, not bad.

2) THEN upload (or download) that file into your software to find all combos that equal sum of skips.

3) Then run output file from equal sum of skips software process through the filters....

The resulting combinations to play would be very significantly decreased, 8,000 or a couple of thousand, or with all the filters, maybe a lot fewer... However, the 6 winning numbers would have to make it through all the filters.

Maybe not so delusional after all?

And how much computing would all this be for the average PC owner? I have equivalent to a Pentium 4 with a half a Gig.

Btw, and Seriously, do you know where I can get a software that will generate a list of all possible combos using 25 numbers?

 

[Babarlish]

Btw, and Seriously, do you know where I can get a software that will generate a list of all possible combos using 25 numbers?

Hello Mirage ,

Isn't that what a wheel does?

Bab

 

[Beaker]

(Originally posted by Snides)


<snip>
Btw, and Seriously, do you know where I can get a software that will generate a list of all possible combos using 25 numbers?

Sorry I'm interrupting but combos of 6 from 25? 177100 is correct

 

[mirage]

Originally quoted by Babarlish

Isn't that what a wheel does?

Well, in fact a really, really big, really comprehensive wheel would be one way to approach this problem, but all combinations of 25 is not a wheel. I don't know what the exact definitions of a wheel would be but I believe the idea of a wheel is to optimize one's chances of winning prizes, small prizes especially, and maybe if you are very lucky, the big one -- by mathmatically designed usage of the xx(numbers), so that you don't have to play all the combinations of that number, as so prohibitively expensive to do so.

As this is one approach to the problem, do you know where one can find really big wheels?

 

[Snides]

- What about the suggestion of filter: include all combos that have skips since last time drawn of 5 or less? I think the average of this should be about 3 numbers - that were last drawn within and including the last 5 draws. Of course includes repeaters and those numbers that were maybe drawn 2 - 4 times in last 5 draws ("hot" numbers) as well as any others.

- You mentioned "position" and "position change" as a filter. What is this?

I think that is a great idea. most of the time that I wheel my tickets I either completely ignore all cold numbers (7, 8, 9 draws back, somewhere in that range) and just wheel my choosen hot/warm numbers. Hmm, crap, that changes a bunch of things.. I just tried running that in the background, and if it only looks at the hot/warm numbers it won't add in any cold, thus your skip sum is much lower, i tried to wheel for a sum skip of 49, but of course it didn't produce any tickets because in order to get a skip sum of 49, all the numbers would have had to come from 7 draws back (which is what i limited it to). So this idea is good, but requires some re-designing.

as for position and position change, well i guess the first one should be position limit.. take the numbers in sorted order, there is a certain limit for each position that is more frequent than others.. Position change has been discussed on here before, sort of the same idea. The numbers in there sorted order will often only change by +-4, with some changing by quite a bit more sometimes, and sometimes there is a pattern of if they went up last time, they go down the next time..

 

[Snides]

Maybe not so delusional after all?

And how much computing would all this be for the average PC owner? I have equivalent to a Pentium 4 with a half a Gig.

Btw, and Seriously, do you know where I can get a software that will generate a list of all possible combos using 25 numbers?

This idea doesn't sound delusional at all.. if there are certain numbers that you don't think will hit, why bother putting them in the set you wheel from? I always do this.. I have different methods that rate numbers into different groups, and quite often several of the groups produce 0 winners so i use them to eliminate numbers to wheel. I would say it's a great idea, and sure you may lose chances at getting the jackpot, but if you still have 4 or 5 numbers in the selected ones to wheel then you'll have a tighter wheel that should produce more small wins, lots of 3# hits, multiple 4# hits, and a better chance at a 5# hit, which is usually at $2000 which would be a decent hit as far as I'm concerned.. some players are jackpot hunters, and some are just looking for supplemental income.. if you can double or better your money each, or many draws, you should remain ahead of the game and have more money in your pocket..

oh ya, you were asking about computing power.. sounds like you have enough for this, though you didn't mention your pentium 4's clock speed, should be in MHz, probably 1.something or higher..
I'm running an AMD 2000+ (which is supposed to be equivalent to a pentium 4 at 2000 MHz) and a half gig of ram as well, and running the program i'm making you takes about 10-12 minutes.. the more filters added and if we start removing numbers from the wheel will speed it up even more.

 

[Snides]

As this is one approach to the problem, do you know where one can find really big wheels?

you could try the program i wrote for wheeling and provided a link to on here awhile back in this thread

you can just click on the 25, or however many numbers you like and wheel them up to get a full wheel, or you can select "use standard filters" and get a filtered wheel using some of the filters we've been talking about..