I did this earlier this morning and something happened it kicked me off so I'll try it again.
This was from the 12-30-2019 draw here. It is the first picks options with no filtering done. It's what it looks like as soon as the draws are updated and calculated. The only editing done to the photo is the headers above each subset has been colored to match the cells so it doesn't give too much away and picks 2-5 are not showing. The picks are laid out though as they are on this games worksheet.
If you take columns B-AA, 3-15, those would classify as first key odd. Cells in B-AA, 18-30, those are first key even. All your options will show up in the non-colored cells only and the headers above each will let you know where they are.
If you see the dark grey columns, those separate the two keys into the 4 subgroups. B-P, 3-15, will be key 1 odd key 2 odd. R-AA, 3-15, is key 1 odd key 2 odd. B-K, 18-30, is key 1 even and key 2 even. M-AA, 18-30, is key 1 even and key 2 odd. (This is not the exact key numbers, but you get the picture this way.) For key 1, if you set it as odd, rows 18-30 will clear off and you will not see them as a pick. If you set key 1 as even, it will clear off rows 3-15 and you will not see those as a pick option. Set key 1 as odd and key 2 as odd, you will only see B-P, 3-15 and all other options will disappear. Set key 1 as odd and key 2 as even, you'll only see R-AA, 3-15 as options. Same with key 1 even, key 2 even and key 1 even and key 2 odd.
Her you only have 3 of the keys to choose from. Odd/odd, odd/even and even/odd. The math on this one made it so that there was no options for keys even/even.
For odd/odd, you only have four choices. 11,5,8 and 9.
Odd/even you only have two choices. 1 and 4.
Even/odd you have four choices. 6,3,2 and 7.
In cell AE3 you see that 6. That lets me know, just by glancing that there is only 6 options for key 1 odd. AE18 let's me know that there is 4 options for key 1 even, and AE16 will give me a total of how many options total per pick. The more stress you can take off the brain, the more relaxed you are to actually look correctly. It is for me anyway. If you throw out 40 numbers per pick and are expecting to pick out of those only 1, your brain can go a little haywire. The simpler the better I've found.
One of these 4 subgroups will have your number in it over 70% of the time. According to my book, which is set up to do a full 31 days at a time, a full month, from January 1, 2017 to December 30, 2019, I'm showing it happens 22 times per month on average. 2019 it happened 267 times and 2017-2018 it happened 269 times each. That's out of 365 games a year. That's less than 100 games a year that you will not have the winning number from the get go. Seems like a lot I guess, but it's part of the purpose of this. To cut out as many options as possible to increase our chances of a win and still give plenty of opportunities per month to get THE WIN. It's plenty when you look at the whole picture.
Back to it. I now know that there are only three ways to set those two keys, with only one being correct. I'm not worrying about if the winning number is in there because I know that the percentages tell me it has a better chance than not to be. I'm only focusing on picking one, again eliminating more stress and allowing clearer thinking. I could just pick one from each set and go with it and still have better chances, but I've got way too much depth in me to do that. That's still just luck in my eyes and that's not what I've put all this work into this for.
Once the pick options are set by percentages that I choose, they go into these pages and into whatever subset that they meet according to the inner math of the balls drawn themselves. They are not just thrown into a set by chance. This is using the inner math of the draws themselves to decide what goes where. If you look at this one close, you will see that there is 12 columns that can be odd/odd keys. There is 7 columns that can be odd/even keys. For the even/even keys, there is 7 subsets and 12 even/odd keys. It's reversed. As I said, there is actually 128 subgroups per option mathematically. I've cut out all the ones per option that have never produced and that only have a very small percentage of the time. This was what was left for pick one. The smallest pick.
The odd thing about this is the 12 and 7, and 7 and 12 options that was left. 7+12=19 which is 1+9=10. 10 is completion and the next step up from the 0-9 whole numbers. Let's look at this as if there was an X going from corner to corner linking odd/odd to even/odd and odd/even to even/even. Odd/odd you have 11,5,8,9. Total those up you have 11+5=16+8=24+9=33. 3+3=6. Go to even/odd and you have 6,3,2,7. 6+3=9+2=11+7=18=1+8=9. 6 and 9 are mathematical even/odd opposites. Odd/odd has 1,4. 1+4=5. In even/even, you have zero numbers. That's 0 numbers. 0 and 5 are mathematical even/odd, positive/negative opposites. This sort of thing happens NOT just once in a while. Quite often. If you look at the date of this draw it was 12 30 2019. That's 3, 3, 3. What is 3 from completion of 10? 7. 7 was the winning number for this date's draws smallest number. Winning numbers was 7,26,30,33,39. That's 7,8,3,6,3 each summed up. 3 is definitely in the date. 8 is also the mathematical even positive of the odd negative 3. Maybe I'm putting more into that than there is, but this is not just a once in a while occurrence that stuff like this happens with this. Another thing that happens is when filtering down. I filter using all "long shots", I'll get either the winning number or what's left will sum up to the winning number. I filter to the best options, it will do the same. It's crazy. Or maybe that's just me also. But it happens every time. I honestly don't know why. Or even how. But I know it does it.
Another thing in this is the matrix. Ball set changes. It goes through that already when we went from 39 to 43. It has not changed how it works whatsoever. It will as of now, accept up to 100 numbers. I had to have this versatile enough for whatever changes made. One cell change to whatever the ball set is is all it takes to adjust. Add more balls or take some away, it takes one second to adjust it. Need more than 100 balls, that can be done in less that 10 minutes. Change all the numbers to pictures or symbols? Give me 30 minutes and a cup of coffee. It will also handle up to 7,000 draws but if it needs more, that's about an hours worth of work, another cup of coffee and a good stretch. Changing options per pick for this layout? Each pick position has one cell on one sheet that has every percentage of what it looks for right there in view. Want less options for pick one? Give that cell a higher percentage. Want more? Give it a lower percentage. Want to change options for 1,3,5 only? Change those cells up or down. Done in seconds. Let it calculate and see what it does. It's easily adaptable to whatever and that's what I wanted with it. Each "play sheet" (1-31) in this adjusts it's self accordingly to what has happened that draw and keeps that data visible on that page so back testing is a breeze. It also adjusts the next sheet accordingly so you're not working off of the wrong data. Running tests on it to find out what happened where has a special spot that the data goes into as to how many hits it's produced per position and how many 1 of 5's, 2 of 5's and so on per month that goes into another book so I can keep track of how it's doing. Every game that's put into it, it will show you the changes as soon as it calculates. It's all in one, see at once spot.
There is a book that tests how many hits per game per month and wins per game per month. It's set up to handle from 2006 and thru now and the next 10 years if needed.
Another book that tests what you see below that shows me where the next winning number showed up next and how many times. There is also a book that does this same thing but it will collect data from all 128 subsets. That book is how I knew what to get rid of for each pick like below and only use what you see instead of showing all 128 subsets per. The main book is 276,000 KB's so it's not a small book. There are 9 other books that is used for filtering that gets their data from the main book then feeds back to the main book. Those books are anywhere from 30,000 kb's up to about 140,000 kb's. Running all 10 of these at once does put a damper on my PC speed and creating has been a pain because of that. Freezes and having to redo was about enough to drive me crazy. I guess that's where the stubbornness played it's part for me. There are also 5 other books that run tests on, well, about anything. Plus many smaller books of tests and back testing. It's definitely been a journey and a half and hours on end of work and thinking.