Alrighty, I've been working on some code for the past few days for a new system idea I had, and I'll post the logistics up here and I'll play it out for a bit to see how she works.
GENERAL:
The general system idea here is a combination of my calculations, and a spin off on the Vtracs system. However, I want to keep the cost low ($5-$10 / day) so it's realistic while hopefully boosting the odds a bit. It's not going to be a millionaire-maker, but will be fun to play out and with some luck (ain't it all luck anway?) it might make a few extra dollars. I'm going to play all the numbers I calculate (max 10 per day) with either a $.50 or $1.00 bet, but you'll read more below (if you don't get bored, heh). If I play my cards right I might be able to get the GF to split tickets wtih me and we will play $5 each/day. ANYWAY....
IDEA:
The idea is to use the Vtracs ideology, but reduce the system. I am shooting for box results for the Play 3, with straights being a nice bonus. All Vtracs I calculate will be reduced as such: 115, 151, 511 = 115, etc. Basically the system converts past results into these reduced Vtracs, looks at when and where each Vtrac hit(s), and calculates which Vtrac is the most "due" to hit. I calculate "due" by the average skips between a Vtrac vs. when it last hit. i.e. if Vtrac 445 averages once every 13 draws and was drawn 10 draws ago, it will be 2 out (2 more draws till it is 'due'), and a Vtrac that averages once every 15 draws and was drawn 5 ago will be 9 out, etc etc. I then take the Vtrac that is the least out and call that the 'expected' Vtrac. Then we calculate all possible combinations of those numbers, not taking order into count. i.e. Vtrac 123 is looking for numbers 051627 (012567 for ease of reading). Now, all combinations would produce 23 tickets, which is $23 or $12.50 / day. Still too much. I decided I wanted to eliminate one of the numbers and come up with a 5 number combination, which would produce only 10 tickets. I included in my calcuations to remove the number that hasn't been seen in the longest time, and it seems to work OK so far. So, if the expected numbers were 012567 (Vtrac 123) and the number of that group that hasn't been seen in the longest is 5, we would be left with the number set of 01267. That produces 10 unique tickets:
012,016,017,026,027,067,126,127,167,267
I included in my calculations (as I usually do) the option for varrying pool sizes (amount of numbers to look at for statistics) as I mention a little bit about in my intro thread. My reporting code loops through all possible pool sizes and totals the hits for that pool and prints out which are most active, which lets me focus on the pool most active and not have to worry that I should be looking deeper, or shallower in the data set.
RESULTS:
Looking at the CT Play 3 (which is where I am focusing) I find the pool size of 73 to be producing the most results, so that will be the pool I use unless otherwise noted..
To decide if this system is worth anything at all, I printed out a report into Excel of how this would have worked since 1/1/2008 and ran a bit of calculation for total wagers, total winnings, overall over/under. With a pool size of 73 this hits 13 out of the last 156 drawings. Nothing to write home about with the basic box payout of $41.50 really, but has some potential (atleast I think). Although, the $.50 bet ($5/day total) pays out $83.50 for a 3-way box (2 numbers the same), so that would increase the winnings a bit. With the 3-way box and 6-way box payouts since Jan 1 the totals would be:
Total Wager: $780. Total Winnings: $707.50. Over/Under: UNDER $72.50
I decided to look at what the damage would be for $10 bets, $5 me and $5 for the GF . While this costs more (double actually), it unlocks the potential for Straight wins which have a much nicer payout.
Sooo, using the same data set but $10 wagers: Total Wager: $1560.00. Total Winnings: $707.50. Over/Under: UNDER $852.50.
BUT, these totals don't take into account any straight hits, which I can't calculate yet since I don't know what my picking strategy would be for ordering the numbers. But we can say that a straight hit would pay out $250 more than the box hits, so 3 striaght hits would put us just under our total wagers while 4 straight hits would put us $150 in the positive! Winnings of $150 from $1500 worth of wagers in 6 months time won't exactly make you strip naked and run down the street, but maybe we can increase this shortly (I need more time to think )
So another brief thought on straight hits.. 4 would put us in the positive, and there were 13 box hits total. So we would need to guess correctly 30.8% of the time.. Since each 3 number ticket has 6 different order combinations, we have 1/6 chance of guessing the straight payout (assuming that ticket were to hit) which is 16.7%. Using these numbers we would probably guess the straight payout 2 out of the 13 hits, but I like to be optimistic and think I'm luckier than that .
So I think that's about all for now. I'll make a post following this one for this system's predictions for tomorrow and we'll start it off!!
Questions, comments? Post 'em up!