Back Testing Formulas

Reply #49

Sarge0202 — Sat, 08 Jul 2017 22:24:39 GMT

(First let me clarify, my programs do NOT include any pick 3 or 4 since the rules are different from the other games)

Yes, I have made errors in my programs before. Some are logic errors and others are flat out math errors. I have been developing these programs on and off for about 20 years and been playing very infrequently. I started off with simple programs basically getting the hottest numbers . One day I decided to test what my profit loss was going to be if I went all the ticket sets. L... [ More ]

Reply #48

Johnebench — Sat, 08 Jul 2017 19:58:10 GMT

Were the Results Really Wrong or did you create a Winning System By Mistake?

A lot of things got created by mistake, Microwave Oven, Penicillin, Ink-jet Printer, Potato Chips, (while working on a rat poison the scientist told the hard of hearing lab tech to TEST IT and he thought he heard taste it and voil , Artificial Sweetener.

Go back And try to re-create it.

Reply #47

AllenB — Sat, 08 Jul 2017 04:47:41 GMT

I think I described this before but bear with me for a minute. In order to Run A test of a System, you first have to teach the Testing Program how to Create the Books . It sometimes takes me awhile to get the System from the information Posted. I go Through a Bunch of Test Runs as the System comes together. A lot of those Tests show Startling Results. That is always an indicator that something is Wrong. Counting thing twice, overlapping formulas with address errors. I am not talking formula erro... [ More ]

Reply #46

AllenB — Sat, 08 Jul 2017 01:40:44 GMT

Just another Vendor Column in The Books .

It's all about keeping the Books Current and getting the right product on the shelf.

Reply #45

Sarge0202 — Fri, 07 Jul 2017 15:32:37 GMT

With adding the 3sum into the back testing, my overall odds of wining the minimum or greater has increased and it is higher then the posted overall odds. I was using Foxpro, but with the latest windows 10 updates, it has broken my backtesting program. I am rewriting it in Java (and learning java), and then I will have to redo all the backtesting. Basically I was using different seed values for a random number generator having it create a set of numbers (with certain checks) and see what % won an... [ More ]

Reply #44

bobby623 — Fri, 07 Jul 2017 14:47:16 GMT

How much have you won with this method?

How do you choose which data sets will provide specific integers that will match the winning combination for a specific drawing?

With that much processing, you must end up with a large number of possibilities that might not be as good as using random selection.

Reply #43

Sarge0202 — Fri, 07 Jul 2017 14:03:33 GMT

I actually use 4 different computers at once when I was doing my back testing. I ended up having to upgrade some power supplies because I would blow them out having max cpu for 12+ hours at a time.

Reply #42

amber123 — Fri, 07 Jul 2017 05:37:00 GMT

Old-School...

Reply #41

AllenB — Fri, 07 Jul 2017 05:08:29 GMT

Or a Bunch of Computers! LOL

Reply #40

amber123 — Fri, 07 Jul 2017 04:39:58 GMT

I-forgot-to-add-that-about-3-months-ago-I-wanted-to-do-a-similar-thing....monitoring-multiple-systems,-and-only-playing-the-ones-which-were-on-a-winning-trend.

Also-....knowing-when-a-trend-is-about-to-go-cold-is-very-important....What-we-need-is-a-super-computer-to-calculate-thousands-of-scenarios.

Reply #39

amber123 — Fri, 07 Jul 2017 04:29:56 GMT

LMAO-with-the-Scotch-tape-store...Ha!

But-if-you-can-build-ALL-those-parameters-into-an-excel,-that-would-be-awesome-and-challenging-to-say-the-least----unless-you're-the-type-who-can-whip-it-up-in-a-jiffy....Like-you...

Wink,-wink...

BTW...The-detail-of-the-explanation-you-made-was-very-well-thought-out...I-commend-you-for-the-illustrations....Not-like-I-need-pictures-but-some-parts-like-the-columns-part..were-very-interesting..

Reply #38

AllenB — Fri, 07 Jul 2017 04:23:32 GMT

There are no numbers in a pile of Dung either. There may not be any numbers in a pile of dung but there is $$$$$$ in it.

Well Said Johnebench. Sound Like you might be enrolled in the College of Knowledge. I Think I am wrapping up my sophomore year, but I attempt a few advanced courses once in a while.

I have learned one important thing and that is to not throw away my money on something that should happen just because it happened before.

I once heard or read some great advice. Don... [ More ]

Reply #37

Johnebench — Fri, 07 Jul 2017 03:36:04 GMT

I back test then I back test the back test looking for something different, Try this try that always looking for that elusive magic formula and just when I think I found one it's too late. It's at the end of it's cycle for who knows for how long. So I put it on the shelf and look for something new/old different that I might have forgot while back testing and start all over again. Now it's all filters of sums, mirrors, flips and flops.

I have learned one important thing and that is to not thro... [ More ]

Reply #36

AllenB — Fri, 07 Jul 2017 02:34:36 GMT

So many of Us running our own Back Tests . There are probably as many Test Setups out there as There are People Testing. It occurred to Me that there should be a Standard way of Doing a Back Test . By Standard, I don't mean My Setup, I mean The List of Questions that a Test Should Answer.

The Obvious First and Only Question Is the System Profitable ?

From what I have seen so Far, there is No System That is Profitable over a long stretch of Games in any State. Most of the Systems produce... [ More ]

Reply #35

AllenB — Wed, 28 Jun 2017 14:54:04 GMT

The overall odds of wining anything in the Illinois lotto (6/52 game) is 13.6%. Using one of my systems I can get as high as a 24.8% overall odds of getting ANY prize.

An increase of 11.2% (13.6% to 24.8%). Sounds like a good system. I am a little confused about the percentages. When you say anything , That sounds like the lowest prize offered for the game. Your data columns are not lined up so I may be wrong; but, it looks like you had 0 (matching 6). 1 (matching 5), 6 (matching 4) and a bun... [ More ]

Reply #34

Sarge0202 — Wed, 28 Jun 2017 13:15:34 GMT

So far it has not been possible to produce any IF statement that hits more than 2% over the expected ODDS.

The overall odds of wining anything in the Illinois lotto (6/52 game) is 13.6%. Using one of my systems I can get as high as a 24.8% overall odds of getting ANY prize.

RANGE MATCH0 MATCH1 MATCH2 MATCH3 MATCH4 MATCH5 MATCH6 MYTOT MYPRCT OVERALL GAMECNT C1 C2 C3 C4 C5 C6

482643 94 100 54 9 1 0 0 -109 57.75193798 24.80620155 258 25 37 48 38 45 46

1259736 105 89 55 8 1 0 0 -11... [ More ]

Reply #33

Oscargrouch05$ — Wed, 28 Jun 2017 11:56:48 GMT

Trigger Doubles Numbers : 05 16 27 38 49

019-715 Trigger Triples Number or Family Members Numbers

Reply #32

AllenB — Wed, 28 Jun 2017 05:50:07 GMT

I am probably going to regret this; but, A bit of the Grape and a whiff of a fire, both grown right here make me a bit Reckless some times. It's a good thing I am happy at home. I can't get into to much trouble here, except maybe HERE.

So here Goes,

Again! This is not, although it could, be meant to Advocate of a new system. Just an example of how a Look Back can be Set-up.

Mostly on the John lately I have had a vision of all of the Pick 3 Games played every day. I am limiting to the ga... [ More ]

Reply #31

amber123 — Wed, 28 Jun 2017 05:32:53 GMT

First watch the video in the hyperlink in blue at the bottom.

Then after watching the video, try to find biases involving consideration of the number 9. Also find what triggers would help you identify a prospective fan.

Reply #30

AllenB — Wed, 28 Jun 2017 04:42:34 GMT

Give me some details and I will give it a Look Back

Reply #29

AllenB — Wed, 28 Jun 2017 04:28:14 GMT

The 'IF' statement is not the only query for a programmer, I will put all queries under 'assumptions', meaning you test the assumptions. I can say the word 'DUE' takes a different angle from 'NEXT' when tracking a parameter(eg sums, digits, deviations, interval points etc).

Relative to my self taught understanding of Excel, If always takes an If Statement to get the 'Due' or 'Next' as the value of a formula or statement. The only other way is a Vertical or Horizontal Lookup, an indirect Addre... [ More ]

Reply #28

adobea78 — Wed, 28 Jun 2017 04:06:38 GMT

The 'IF' statement is not the only query for a programmer, I will put all queries under 'assumptions', meaning you test the assumptions. I can say the word 'DUE' takes a different angle from 'NEXT' when tracking a parameter(eg sums, digits, deviations, interval points etc).

So far it has not been possible to produce any IF statement that hits more than 2% over the expected ODDS.

Unless you change picks every draw, a strategic wager is more concerned about the 'PRIZE RATIO' not 'The number... [ More ]

Reply #27

AllenB — Wed, 28 Jun 2017 03:43:40 GMT

The following is a summary of a 500 Game Test run on every State. In this Test Each State uses itself as the Base. Meaning the formulas are applied to the previous draw of the same game. Each Test is based on playing the formula results in the Next Game only. The Table shows the Offspring with the most hits during the test. The Maximum hit count is in column 4, the adjacent column show how many of those maximums occurred and the last column show the formula that had the most hits. Currently if t... [ More ]

Reply #26

AllenB — Wed, 28 Jun 2017 02:45:53 GMT

If excel is to formulate algorithms in the realm of randomness, then good luck !

That it not what the Formula Test is for. It just looks back as far as you want to see how many times and when the Formula hit. That's it. Period. It is just a look at reality in the post RNG Draw. It takes another Skill Set to interpret the results and make a successful prediction. Unfortunately what it show is that Formulas Hit so infrequently that I don't see how to get a good prediction, YET! I do not want to... [ More ]

Reply #25

Soledad — Tue, 27 Jun 2017 18:38:03 GMT

You'd be hard pressed to find anyone calling you dumb or stupid. Lol

Reply #24

adobea78 — Tue, 27 Jun 2017 18:34:10 GMT

I have said this before in another Forum. It is not the System or about Playing the Game, It's all about the Excel!

If excel is to formulate algorithms in the realm of randomness, then good luck !

Every cycle starts anew, every draw is unique

Reply #23

amber123 — Tue, 27 Jun 2017 04:15:56 GMT

Hi Girlie, nice seeing you reply here.

Glad to see you seeing what I'm seeing.lol..AllenB is like a gift from Heaven.

Reply #22

amber123 — Tue, 27 Jun 2017 04:12:12 GMT

Interesting read Allen on your last entry.

For 30 plus years I've known about the crazy number 9, but recently I've discovered new info that the video below (blue hyperlink) shows, hence my Tesla 369 system. Always trying to tweak those numbers, or any combination of numbers that equal sums of 9.

If you're interested and have the time, please tell me If there's a way you can take advantage of this.

I'm going to spend some time between jobs (down time) tomorrow to find some way to track... [ More ]

Reply #21

amber123 — Tue, 27 Jun 2017 04:00:26 GMT

RJOH means well, been here at LP from almost from the beginning. But he uses that same reasoning every time the subject surfaces. It's all good, I just wish he would respond and tell me what part I wrote was wrong, I want to know. Learning new info is like gold to me.

Unfortunately he never does reply back, oh well.

Reply #20

Ladyc743 — Tue, 27 Jun 2017 03:03:09 GMT

amber123......LOL

Reply #19

Ladyc743 — Tue, 27 Jun 2017 02:58:42 GMT

Totally agree Allen and try to deal with Texas with 4 darn draws OMG is a BRAIN WRECKER

Reply #18

AllenB — Tue, 27 Jun 2017 02:44:04 GMT

Here is an example of a WHEN-THEN test that I Started a few days ago.

I am not floating an idea for a system. It is rather an example of Looking back at what happened next after a particular criteria value happened in a draw or a bunch of draws. Here is an absurd example of what one might want to know. What if you wanted to know what digit hit most often in position 1 after the Sum of the Previous draw was 14? Bigger yet, what hit most often after any of sums 0-27. It is pretty easy to build... [ More ]

Reply #17

RJOh — Mon, 26 Jun 2017 22:05:56 GMT

For example my Classic Lotto file has 1632 drawings and I think I've discovered a pattern that makes it easier to predict at least 10 combinations of which one will match 4 or better. To test it, I create a test file of the first 1600 drawings using only the data in the test file. I do similar tests with other test files, all less than 1632 drawings to see if any have a winner, if not then I have nothing.

Reply #16

Johnebench — Mon, 26 Jun 2017 19:10:51 GMT

This is a tool for back testing systems, It will save a lot of hand writing and paper for sure.

Reply #15

lottolaughs — Mon, 26 Jun 2017 18:36:08 GMT

Another thing....how can you ever test forward until the draws come in? I like to see what's shown in the past and how a system's performed. Call me dumb, call me stupid but I don't want to wait years, starting today, to see how a system I wrote yesterday will perform in the future. Who's got that kind of time

Reply #14

AllenB — Mon, 26 Jun 2017 18:29:43 GMT

Going back and only using the data you knew before a drawing would be closer to back testing but that's harder so it's not done.

The Data you knew before the drawing? I have run a few ideas in the past. I am curious about what data you know before a drawing. That known data for the next draw can also be recorded for all of the previous draws (assuming it can be calculated). You can then look back for the Same known and record what happened next. Call it a WHEN-THEN test.

Probably st... [ More ]

Reply #13

amber123 — Mon, 26 Jun 2017 18:15:41 GMT

I agree, when you go back and test for combinations you already know will come up, it's not back testing

But you don't know what the combinations will come up because you didn't have that system back then. That's the whole point of back testing. You want to see if the system would've been successful if you had the system, let's say a year or two ago.

Reply #12

RJOh — Mon, 26 Jun 2017 18:06:38 GMT

I agree, when you go back and test for combinations you already know will come up, it's not back testing. Going back and only using the data you knew before a drawing would be closer to back testing but that's harder so it's not done. All you confirm is what has happened in past drawings which is no guarantee that it will happen again in the next drawing.

Reply #11

AllenB — Mon, 26 Jun 2017 17:33:59 GMT

Bob, I agree with everything you say. You might be confused about my intent. I am not advocating any System . I only back test them. I can tell you when the System had a hit all the way back to the beginning of a game. I can report on any length of Back test and calculate the cost to play for the hits.

This is just a tool. A tool that generally shows that a System is going to lose. Tests at best show you that there is no quick hit to be spotted from the data. Instead they show that over long... [ More ]

Reply #10

bobby623 — Mon, 26 Jun 2017 16:31:12 GMT

Sorry for the strong language. I spoke before I had my morning coffee!

I'm an old man, too. Been around a lot longer than I thought I would be.

But, my Mom would be embarrassed. She would probably threaten to wash my mouth out with soap!

Still, if you or anyone you know can play the lottery without 'looking back' for clues that would help them pick winning numbers has got to be rich.

Rich folks probably don't visit LP for help in choosing winning numbers. They already know.

If the... [ More ]

Reply #9

Soledad — Mon, 26 Jun 2017 13:53:14 GMT

Intuition my dear good friend sir. In the key of 'simple'.

Reply #8

AllenB — Mon, 26 Jun 2017 13:34:55 GMT

Sorry Bob. I forgot how easy it is to set you off. I am an old man and I can be a little bitter myself; but, You might want to try to remain calmer and keep your labels secret. We all wannabe something Bob.

I take a slightly different approach. Back Testing should be called and is Really nothing more than a Look Back. You take an idea and Look Back to see how it performed over time. The View and the Memory of the View make a few of us feel a little better, Bob. I will admit that the View usu... [ More ]

Reply #7

bobby623 — Mon, 26 Jun 2017 12:35:00 GMT

Back testing is a pile of stinking crap being put out by wannabes.

The human mind is a wonderful tool, but if we know the outcome, how do we know we aren't just repeating something we already know??

I seriously doubt that anyone has a system that knows what the winning combination was and can tell us why we aren't able to choose winning integers.

We analyze, we choose, we win or lose and then move on to the next drawing.

We don't need some system to tell us where we went wrong.

Ca... [ More ]

Reply #6

BabyBoy33 — Sun, 25 Jun 2017 17:42:10 GMT

Can Someone Show Me How To Do A Back Testing Formulas For Texas

Reply #5

amber123 — Sun, 25 Jun 2017 16:40:58 GMT

I agree, this is such a time saver. Where was Allen 8 years ago?

Great work Allen

Reply #4

lottolaughs — Sun, 25 Jun 2017 12:36:54 GMT

I just think this is so incredible! You are a gem

Reply #3

AllenB — Sat, 24 Jun 2017 19:28:02 GMT

There are also WAY to many options. Comparing Every State, even though a macro does the work in a couple of minutes, is a giant block of data to wade through. Extracting the Top performers as I said previously only show you what Might Have Been and is no guarantee of the Next.

When to Play is also Key. An Off Spring Column (one of your formulas) May show more hits if you waite 2 or 3 drawings to play. How Long to play is also Key. If I find that a column also hits if I waite 1 Game, Playing i... [ More ]

Reply #2

AllenB — Sat, 24 Jun 2017 19:12:27 GMT

I appreciate your enthusiasm. I am not so quick to call something Junk. It only shows how a formula has behaved over a specified number of games. Small Game Blocks with more than 1 Hit show success, Unfortunately that might be it for another 100 Games. since the best performance so far is 4 hits in 500 games, 3 hits in 300 games finding success means you are going to have to waite a while for another hit.

Reply #1

lottolaughs — Sat, 24 Jun 2017 18:21:34 GMT

So,AllenB,if I am understanding you correctly, you've come up with a program that can automatically back test our..ahem..systems,lol. That is amazing. We'll be able to know right away whether our system is junk or not. I love it! Think of all the time we'll save. Priceless

Back Testing Formulas

AllenB — Sat, 24 Jun 2017 16:35:26 GMT

Many of the Ideas for Systems posted here are formulas applied to the last Draw. A few examples posted recently are LottoLaugh s Runaways and Amber123 s Tesla_Pick-3.

Any of these Formulas can be put into a Off Spring Table. LottoLaugh s System has 18,000 Off Spring Combinations (18 for each of the 1000 possible Draw Combinations), while Amber s Tesla System contains 17,000 Off Spring Combinations (17 per 1000).

Either of these Tables can be used to produce the combinations In Play for t... [ More ]