hypersoniq's Blog

harsh on the win column so far... LRS only got1 hit in pick3, droppedit but still update it and it will likely be the only hit LRS gets this year.

Revisited my vertical pair concept from days gone by for pick 3... no way to backtest still, so no plans on moving it forward, it was moreof a coding exercise to streamline the formulas and speed uprecalcs... that part worked.

powerball flatlined for me since that p3 hit... not even a 0+1 in MONTHS...

gotta find a new road to travel for that red ball, what I got ain't cutting it.

NOT the first time it hashappened... the red balls are poised for a 5-ball palindrome

last 4red balls in order ... 26-35-20-35-?

I'm playing 26 tonight!

my system calls for a 02 to come up, so I will break protocol and play a second ticket tonight (Qp the white balls and add a 26 as the RB)

I think this could be the first one of the new matrix, the old matrix had at least 2 and a few close calls.

If anyone recalls the old matrix, a few years back I noticed that 20 follows 35 and 35 follows 20, but usually with 5 to 11 draws between them, this was a first I believe.... that they were back-to-back-to-back!

this may be nothing, but I am hoping it is something!

playing that pick 3 "matrix" lookup table excel sheet is an interesting pasttime, but I need to optimize and automate this endeavour.

the first time i used it for 1-1-1 replacement I managed a backtest of 20 hits on +1,+5,+0

then I tweaked the numbers and ended up with 55 hits.

I used the system to gather follower data (what number follows each number in each position the most) this only produced 19 backtest hits.

tweaking those numbers and it's up to 59 hits... I can't break the 60 hit "barrier".

the basis of the system is using last night's draw, then replacing the numbers with the numbers on the lookup table... changing the table numbers affects the backtest. allows you to "dial-in" more hits, but like I thought, the most frequent numbers do NOT come out together.

I need to figure out a VB program to try many different scenarios, but I can't get a clear picture of the logic required...

I need to figure this out before attempting the same thing with powerball...

59 hits averages 2 hits a year, the system would have paid for itself AND brought in $19,000 profit...BUT it would have taken 26 years and missed 5 years... 2 in a row

based on backtest results, it may miss this year so I am not playing it, but I AM tracking it... could be interesting to try with powerball if I get the infernal logic figured out.

BEFORE THE WIN 

1.) Create personal property trustS (note the plural). The reason there are two is that I will claim the prize myself, but as a trustee. the name will imply multiple winners, but that won't be the case ;-) The SECOND trust is to transfer ALL of the assets into after the electronic transfer goes thru and I have access to the money... WHY? asset protection, the second trust name will NOT be publicised, therefore no frivelous lawsuits against it.

I am in the process of getting those ready NOW with my trust-authoring software... they will be ready to notarize in the event of a win, NO attorney required!

2.) print out the map to lottery hq for an easy trip in what I am guessing will be a frenzied state of mind. (DONE)

AFTER THE WIN

1.) Sign the ticket with my name AND the first trust's name, my bank already handles CDARS, so I just need to get their electronic transfer number before claiming the ticket and up to 30,000,000 will be insured in a CDARS account.

2.) Scan front and back, original goes in the safety deposit box until the first business day I can make the trip.

3.) DECLINE PRESS CONFERENCE!!! DECLINE PHOTOS!!! Name will still be released, but I have half a chance this way of my clean escape from Pennsylvania!

4.) Take the CASH OPTION so I can take advantage of PA's NO TAX benefit on the lump sum. 

5.) Settle all debt ASAP, no strings left to tie us to this state.

6.) Identify a CDARS bank where we will move (Done), transfer funds there next time we go for vacation (scheduled for june '08... sooner with a win)

7.) BLACK OPS move... no forwarding phone/address/cell phones/anything.

8.) Open my own business (already know what, but I will have the startup funds!)

9.) keep bulk of the money in the trust (second one) to protect the balance. Diversify a portfolio with a large chunk of it.... gotta keep an INCOME stream open...

10.) Pay for the new home IN FULL immediately, earmark a portion of the winnings to sustain that lifestyle for 10 years (calculate inflation and include personal funds... gotta pay yourself too!)... this means no mansion, but who cares... My dream home has no minimum square footage and will be as close to the beach as We can afford within the planned budget.

11.) Cruise the LP on the Wi-Fi alienware laptop poolside while watching the waves roll in!

miss anything? got a plan of your own?

...with NO picture on the box!

still trying to figure out what to do next...

maybe look at what I have done over the last few years and see what can be put together...???

take some old stats and use them to guide the "matrix" game...

more with deltas and "positive change"...

LRS and RNG together for pick3 mid?

who knows...

anyway

happy July4th fellow Americans, and have a great day wherever you happen to be from!

Was toying around with the "Positive Change" thing... best I could find was 20 hits in 10,400+ draws. BUT I used a lookup table rather than a formula... Started running each position thru it's paces and found I was able to tweak the wins to 55! that's major profit over 27 years with only 5 years missed (including 2007 so far).

I think 55 is all i will find... but I would like to try this for other games... replace numbers while looking at the results... some numbers are cut altogether and some numbers are doubled up... system has not been hit since jan. of 2006... so it's "waiting year" is up and it's DUE.

no advanced forecasting on this one since it works with the last draw only...

I will be saving money this month... only playing cash5 when it gets to 200,000 so that will save at least 15 bucks for july. Also dropping the match6 until it rolls to 5 million (then I won't play it 'till at least october if it ain't hit)

I still need to update the c5 sheets and check the website to see the next jackpot to make the go/no-go decision.

powerball remains unchanged... one ticket per draw with the power play... but I will be trying that lookup trick with the red ball to see if it's feasible

Also picked up a sweet deal on a program called "Perfect Attorney"(Platinum edition) which has tutorials and form generation so I can whip up a trust document and be ready should lightning strike.

I have decided to claim it myself if I win, but also use a trust as a protection device. I will take my 15 minutes of fame to advertise the skatepark i will build... can't BUY that kind of publicity! and the trust will prevent me getting sued running the skatepark (AND keep it safe from any type of lawsuit my wife may encounter as an R.N.)

I finally figured out the last piece of the puzzle... now it's time to win!

$66 to play... $1 won to date (Cash5)

Not planning a July test since I'm back on my own playing money again.

LRS is ok, only one that won me money so far. Will stick with LRS and the current matrix for powerball, but that's it for the other PA games for me for awhile.

One last cash5 ticket for tonight, will need a 4/5 to make july a success... not counting on it.

Powerball has been a big goose egg for all the tests this year so far... if it wasn't for the big money (even after it's hit), I would drop that one FIRST.

I will still be tweaking the other systems for paper play (mostly the pick3) but pick3/4 lost last month and were axed... Match6 went bust and will be cut for july and the cash 5 has one last chance to save itself @ 7pm tonight.

otherwise I am back to just PB... one ticket per draw with the powerplay.

other than LRS and the 1-1-1 system I am tweaking and trying to automate, I am out of ideas for the moment.

As always, the easiest way to start out atheory is against pick 3 data.

I had previously worked with the "1-1-1" system. I had performed operations in excel that subtracted 10 and used absolute values of negative numbers... too complex!

I have redone a "proof of concept" worksheet and have decided to adopt a rolling lookup table. there ARE no negative numbers OR numbers over 9.

problem is it has to be manually changed, there are 1,000 possibilities! I got thru several dozen so far and realize this will have to be automated.

so since nothing is subtracted and just rolls forward, the name "positive change" comes into play (well, the positive part...)

The change part is the new "all positive" delta. using a rolling lookup instead of formulas that manipulate numbers has one major benefit... lightning fast re-calculations!

I need to figure out how to do this in VB! Guess I should search for character replacement algorithms...

watch the pa pick3 tonight... if 6-2-8 is the evening pick3 number, then this is another lucky coincidental system!

Planning a big test on the Pick3 PA eve. AND Midday (NOT a live playing test). THis is me vs. history. Going to revisit the 1-1-1 system, for all 1,000 combos from

0-0-0 to 9-9-9 (Where you add 0,1,2,3,4,5,6,7,8 or 9 to the previous night's number)

using the "rolling 10" ... 8-9-0-1-2-3-4-5-6-7 etc...

Once I have the data, I will attempt to program this test in VB and compare the manually obtained results...

THEN apply it to the numbers AFTER LRS was run.

Remember, in actual practice, LRS was the only system I had work for real money, but 1-1-1 backtested TWICE the hits last time I ran it.

I need to start looking at what's next... LRS is a great tool, but there has to be more to pick3 life than a few hits every decade... 

Check the blog for further updates...

My own research lab

HyperSoniQ Innovations

and the first order of business is research into ideas I have about alternative energy vehicles, and alternative energy in general.

This would be the springboard for every idea I have had that would require research and development to get to market (or just to get to a patent which can be licenced to a manufacturer).

Once a decent management team is in place at the skatepark, here is where I would spend the majority of my time that I currently have to waste working for someone else.

Come ON MUSL, I want to see the press dilute the lab's contributions with gem headlines like "Lottery Winner perfects next generation storage cell technology for Electric Vehicles"...

I would open a huge Skateboard and BMX bicycle action park.

I would incorporate Ideas I have drawn up for the ramps and flooring that would allow them to ride like normal but if they fall and impact, it would reduce the chance of broken bones.

I would create areas that would match rider's skills and budget...

Public area($3 all, $2 straight B's, $1 Straight A's)... I need to win the PB to keep this admission artificially LOW so more people get the opportunity to participate. There would be plenty of challenging obstacles but done safely, not like those public concrete pool parks. Here I would work with local school districts... ONE by getting kids to bring in copies of their report cards with discounts for all A's and B's, and TWO to get kids in high school who need to use community service as part of their terms of graduation to teach others basic tricks. That both helps the teacher AND the others.

Training area($5 additional, $4 straight B's, $3 Straight A's)... This will cost more to get into but will have PAID instructors. Here riders can learn more difficult aspects of riding like big Vert ramps or Backflips into foam pits (for safety, and it's just fun). Also a unique indoor BMX dirt racing track for practice in any weather.

Core area($2 additional, $1 straight B's, FREE Straight A's)... This will be the X-games sized ramps, park courses and dirt jumps... Visitors will need to be "checked" by instructors to qualify to enter (for safety). I would line up pros to do clinics and have pro-am competitions in here. No amount of money will get you into the core area if you are not signed off as capable by a staff instructor.

So access to the core area would be $10 max ($4 if you have straight A's)

on the cheap, a straight A student could ride for $1 a day.

these are just estimates... but I couldn't hold these kind of prices if I wasn't already a millionaire...

The whole park can be built in phases, with Public being phase one. I am looking into doing this anyway once we head south next year... but I would rather not have a huge loan and high admission prices...

come on MUSL... it's a good idea!

Some seeds took as long as 4,000 loops to pick their number, some took 7, several took 1 (The same 9 from the last test).

there are only 1,000 combos... so there is a definite problem. The code is right on the money, I am just wrong with my seed assumption (whole date number, INT process)

I have 3 more variations...

Date+0.5,INT

Date,Round

Date+0.5,Round

I am currently running Date+0.5,INT... Still lets me browse, a definite improvement over it's ancestor the "Coincidizer" which held this PC hostage for upwards of 11 hours... This takes about 60 minutes at normal priority. ONLY because some numbers require several hundred and several THOUSAND loop iterations.

I already have VB's RANDOMIZE sensitivity threshhold identified at 0.5, and I have seen a marked improvement over LRS versus this CPU generated game...

looking for about 200 hits of the 1597 draw history before I start playing for real.... A number I am NOT likely to see considering so far the best is 9...

and no apparent practice draw patterns have emerged.

Unfazed by the initial noise, I move onward... (hopefully) to Victory!

had a cash5 ticket where LRS called one number... three of the other four were ONE OFF. all higher.

I am adding a tweak tothe LRS formula to check and count this phenomenon as one more fine-tuning step for LRS, particularly noticeable on the cash5.

rather than re-run a sheet copy I am just going to add to the LRS formula line to add the contents of designated cells and fill them with numbers from -5 to +5 so I can count the correct hits of each modification.

I caught and fixed a logic error in my new CR code, so hopefully I can update and run tomorrow.

Need to code a new button to run the test.

This one will make several assumptions...

1. The integer value of the DATE is part of the seed

2. They do pre-draw iterations of the loop

This will be a time-consuming run. One variable picks how many times it will iterate... going to start with 10

For each number in the history, it will randomize based on that date's integer value and run thru 10 iterations (each iteration is just one step on the list)

output will be the 10 RNG numbers, a "Yes/No" indication of a match, the number of the iteration that matched (1 thru 10) and the seed value that was used.

round 2 will change the variable from 10 to 100 and eliminate the output of picked numbers (to speed run time).

I said it before, it probably CANNOT be done, but if it can, I WILL DO IT THIS YEAR!

LRS continues, Cash5 went cold... hoping weekly re-scan changes things a bit... powerball has flatlined on LRS and Match6 is suffering a severe case of "one-off-itis"

Switched from full history to new matrix only in the PB... First play 2nite!

Looking for an 11 in the red ball slot... THIS may be the jump start LRS needed for powerball.

Test one involved seeding RANDOMIZE in a loop of the numbers 1 to 1,000,000 (one at a time)... generating a random list of numbers equal in length to the draw history... comparing for matches, then printing out the highest ones found.

UNFORTUNATELY there was no great news.

There are 1,570 draws so far on PA mid-day pick3 and the best list only matched 9. YES technically this beats the game on paper (cost to play taken from 9 straight hits = a profit of $2,930) but it is NOT what I was hoping for.

It is better than LRS, which picked 2 straight hits for the mid-day...

but I must move on.

trying different seeds next, ROUND vs INT for the pick, decimals etc...

I am looking for 227 hits, that averages one a week.

So I did show good results if it were a ball drawn game, but BAD results trying to synch 2 RNGs... Ideal would be of course 1,570 matches...

not there yet, but NOT giving up either, the potential returns are too great!

still haven't spent a dime so it's all good.