Of Cat Washing and Backtesting

Reply #29

Delta Draw — Sat, 28 May 2011 03:33:32 GMT

A game like Keno is wonderful to back test in.

But the back testing I am speaking of is the back testing of combination

wheels against a history. The key is to not to try to get a jackpot, but work toward

lesser matches and a lesser payout. You can determine what numbers and

combinations are due and play accordingly. Since most keno games pick 20

numbers, a match 6,7,8,9 or ten play can deliver steady wins. It all depends on

how hard you want to work.

The probability is one

Reply #28

jimmy4164 — Thu, 26 May 2011 02:31:26 GMT

Since noone has yet reported, I guess I can assume there have been no attempts to simulate time*treat's roulette example. I decided to do it myself. Just in case you tried it and are getting tired waiting for a run that results in 11 or more wins out of 20 tries, think about this. You are probably using the probability of winning with a bet on Red by chance, which is 47.37%. Now, ask yourself - would Bluejay set up a challenge where someone with a system that is no better than chance could beat... [ More ]

Reply #27

dr san — Wed, 25 May 2011 21:57:25 GMT

the law of the fragments retractable Witwicky

Reply #26

jimmy4164 — Wed, 25 May 2011 19:11:26 GMT

You're still missing the point time*treat. Of course I don't expect more than 50% of 20 runs of 200,000 spins each to turn a profit. I trust that Bluejay has determined the odds of that occurring are low enough that he feels quite secure with his 10/1 Challenge.

Please accept my suggestion and run the Basic program modified to make 20 runs of 200,000 spins each with a $5000 starting bankroll and see what happens.

You didn't address it here but I hope you see that a system that could eek ou... [ More ]

Reply #25

time*treat — Wed, 25 May 2011 18:50:15 GMT

jimmy,

The reason for showing the calucation as a series of 3 extremely long winning and losing streaks was so that people wouldn't get bogged down trying to follow a bunch of plus and minus $5s.

I did point out that a system would need to win at close to 1/2 the time: 49.75%. I also said this was a worst case success rate.

The real question is not whether a $5,000 bankroll could ever survive such a play method, but whether it would do so 11 times in a given string of 20 trials. If you... [ More ]

Reply #24

jimmy4164 — Wed, 25 May 2011 15:20:56 GMT

time*treat,

You've performed some perfectly reasonable and correct calculations. However, you've used them to draw invalid conclusions.

In order to show a profit of just $10 after 200,000 spins, your system has to generate a sum of money equal to your bankroll X 100 -- as a worst case scenario.

You're forgetting something very important. Because of the nature of this game, pure randomness alone would allow you to win [close to] 1/2 the time, and consequently, your $5000 initial bankro... [ More ]

Reply #23

time*treat — Wed, 25 May 2011 04:28:38 GMT

I wanted to see what the minimum performance needed to be for a simple $5-always-bet-on-red roulette system to survive 200,000 rounds and still show a gain. The $5,000 bankroll will take 1000 in-a-row losses, I put those at the end.

That gives 199,000 rounds to play with. Because the number of bets is even, you need 2 more wins than losses (of the 199,000 spins), for a net gain.

starting bankroll: $5,000

rules: bet $5 on red, each time

winning streak: 99,501 reds in a row (99,501

Reply #22

Stack47 — Tue, 24 May 2011 18:09:40 GMT

A system needs a 20,000.02% profit to win a dollar.

Do you get it yet or do you suffer from innumeracy

Reply #21

jimmy4164 — Tue, 24 May 2011 14:39:20 GMT

Are you sure about that?

Let's go to the source, with a quote from the Challenge:

But no one plays 200,000 rounds!

Some folks have said the challenge is unfair because a system could win in the short term,even if it can't last 200,000 rounds. They're missing two things. First, I'm offering a whopping 10 to 1 odds. With such generous odds, even a system with a lousy 11% chance of beating the challenge is a good bet for the challenger (and a bad bet for me and my $30,000). Their system... [ More ]

Reply #20

Stack47 — Mon, 23 May 2011 11:43:48 GMT

I was aware of that many years before ever posting at this website.

How many more years will it take for you to figure out Bluejay's challenge is a parody of gaming systems

Reply #19

jimmy4164 — Mon, 23 May 2011 00:01:57 GMT

When are you going to figure out systems players are trying to win money in the next few drawings, not in the next 200,000 drawings?

I was aware of that many years before ever posting at this website. As Delta Draw pointed out quite well in the Innumeracy thread, the only cure for what you are afflicted with is education. Sorry, but I don't think I have the time it would require to get you started on the road to recovery. Besides, based on the quoted question, your case may be incurable.... [ More ]

Reply #18

Stack47 — Sun, 22 May 2011 20:22:56 GMT

Your second paragraph is absolutely right.

Of course it's correct unless you play in a casino that forces you play until you lose all your money. The rest of us play in casinos where we can quit at any time so why are you claiming Bluejay's challenge is based on real casino play when it's not?

Do you think this would be a fair setup for you?

It's the same unfair setup Bluejay is using to say Not one system player has accepted my challenge .

Reply #17

Stack47 — Sun, 22 May 2011 20:09:05 GMT

I could live with the 200,000 rounds at a minimum $5 bet per round if the starting bankroll was $1,000,000. But since the bankroll is limited to $5000 the system needs a 20,000.02% profit just to win a dollar. Most players would be overjoyed if the could win $10,000 in 100 rounds with a starting bankroll of $5000.

Without a quit when you're ahead option, you'll eventually succumb to the long-run odds/house edge/payouts mismatch that goes with such games.

Bluejay and Jimmy keep saying the... [ More ]

Reply #16

Stack47 — Sun, 22 May 2011 19:55:39 GMT

Suppose you put up $30K to my $10K in a simulated Powerball game where you bet me that MY system can't beat the lottery over 200,000 draws with a $5000 stake, but with the option for me to quit when I'm ahead, if I choose.

I'd expect that from a 40-year-old just waking up from a 30 year coma, but not from someone who should have learned something just by logging into Lottery Post. PB has 104 drawings a year and you believe 1923 years is reasonable tests. That doesn't even pass the eyes-rolli... [ More ]

Reply #15

jimmy4164 — Sun, 22 May 2011 04:19:49 GMT

Your second paragraph is absolutely right.

Why would you call this requirement a catch?

Suppose you put up $30K to my $10K in a simulated Powerball game where you bet me that MY system can't beat the lottery over 200,000 draws with a $5000 stake, but with the option for me to quit when I'm ahead, if I choose. Let's suppose further that I'm very lucky and hit a 5+0 and win $200,000 on, say, the 3rd Draw. I then elect to exercise my quit option, claim my system beats the lottery, and reques... [ More ]

Reply #14

time*treat — Sun, 22 May 2011 03:40:39 GMT

The catch is in the 200,000 rounds requirement.

Without a quit when you're ahead option, you'll eventually succumb to the long-run odds/house edge/payouts mismatch that goes with such games.

Reply #13

jimmy4164 — Sat, 21 May 2011 19:48:16 GMT

If you ever figure this out, you are going to wish you could erase your posts in this thread.

But I suspect there's no need for you to worry about that.

Reply #12

Stack47 — Sat, 21 May 2011 19:16:35 GMT

Earlier you totally confused the odds by talking about needing 20,000% to win.

Yep, I did because I forgot the system must win at least $1 to win the 200,000 round challenge, but I corrected my mistake by saying the system must show a 20,000.02% profit to win the challenge. You can fill 100 pages with links but it still won't change the fact to last 200,000 rounds at a minimum bet of $5 per round, at least $1 million in bets must be made on a starting bankroll of only $5000.

Do you have p... [ More ]

Reply #11

jimmy4164 — Sat, 21 May 2011 05:32:31 GMT

If you ever figure this out, you are going to wish you could erase your posts in this thread.

The fact that you can't see how liberal Bluejay's offer is, is telling. The house edge in the casino games he's dealing with is slim, compared to the 50% of the lotteries. Playing these games is more akin to playing Pick-3 with a $950 payout for a straight win. As I tried to show you in another thread, with graphs and all, during short terms of play, the house will be beaten, by chance. It's over the... [ More ]

Reply #10

Stack47 — Sat, 21 May 2011 03:24:14 GMT

Jimmy,

Are you still hunting for ocean front property in Arizona?

Bluejay: Length. Each simulation will run for 200,000 rounds. A round is an event where any bet is resolved. It's fine for your system to sit out sometimes to wait for streaks or other conditions, but 200,000 rounds must be wagered on (unless you go bust first). We'll run this simulation 20 times, and you win the challenge if your system wins at least 11 out of the 20 times.

AND

Bluejay: I will program a computer simu... [ More ]

Reply #9

jimmy4164 — Fri, 20 May 2011 21:01:19 GMT

Stack47,

I think you need to put on your critical thinking cap and read the rules again, slowly.

--Jimmy4164

(From Bluejay's Rules:)

Length. Each simulation will run for 200,000 rounds. A round is an event where any bet is resolved. It's fine for your system to sit out sometimes to wait for streaks or other conditions, but 200,000 rounds must be wagered on (unless you go bust first). We'll run this simulation 20 times, and you win the challenge if your system wins at least 11 out of... [ More ]

Reply #8

Stack47 — Fri, 20 May 2011 20:47:53 GMT

Oh, I did read the terms of Bluejay's challenge.

Bluejay: that your betting system cannot beat a game of roulette, baccarat, or craps, as the player, using common Vegas rules, starting with a generous $5000 bankroll, in a computer simulation, as per the additional terms below.

Bluejay: Length. Each simulation will run for 200,000 rounds.

Bluejay: The table limits are $5 minimum and $5000 maximum.

Starting bankroll: $5000

Minimum bet: $5

Length: 200,000 rounds

Bluejay's ch... [ More ]

Reply #7

jimmy4164 — Fri, 20 May 2011 16:13:54 GMT

Anyone trying to beat the Bluejay challenge; all you have to do is find a system that makes a 20,000% profit in 200,000 trials.

If you actually believe this statement, then you have a bigger problem than I thought understanding Probability and Statistics and the relationship between the two. It's patently false! I think you'd better take another look at Bluejay's Challenge. By the way, he replies to emails; why not share your thoughts with him?

Bluejay's $30,000 Challenge

Here's anothe... [ More ]

Reply #6

Stack47 — Fri, 20 May 2011 00:29:13 GMT

As I've said before his test seems to be picking a combination and seeing how many times it would have hit in the past which is the same as calculation the odds of it hitting any time.

I've been hoping we see a test that resembles actual pick-3 or any lottery game play. The only tests I saw were 5000 to 50,000 random players betting $1 every night for years.

But for testing systems designed for pick5 and jackpot games which probably avoid combinations that have come up in the past, back... [ More ]

Reply #5

RJOh — Wed, 18 May 2011 03:13:19 GMT

Back testing and simulation is based on the much heralded Bluejay challenge no system player will take .

That's part of the problem if your definition of back testing is based on jimmy4164 thread. As I've said before his test seems to be picking a combination and seeing how many times it would have hit in the past which is the same as calculation the odds of it hitting any time. That may be fine for testing pick3 systems which only have a 1000 or 220 possible outcomes and over 700 drawings ye... [ More ]

Reply #4

Stack47 — Tue, 17 May 2011 21:49:31 GMT

Back testing and simulation is based on the much heralded Bluejay challenge no system player will take . Can't test a ten combo pick-3 system against simulated play for 100 drawings but no problem believing 50,000 Pick-3 QP players would wager $1 a drawing for 20,000 drawings.

For comparison, take a system's player playing $5 a night/$30 a week for box hits.

Or how about Rcb using a version of Power Trails based on what was drawn last month. How either method would do 30 years ago or how... [ More ]

Reply #3

garyo1954 — Sun, 15 May 2011 17:31:31 GMT

BobP has a lot of very good material stemming from the many years he has spent writing about lotteries and I'd be surprised if he didn't recommend some testing before using it. Tests like that are contemplation testing. It provides a chance to get a feel for how the method works and to get accustomed to using it. Yes. I do tests like that as well.

Its test like RJOH points out that are worthless. If you can count on one hand how many times 999 came out in the history of the Texas night game,... [ More ]

Reply #2

rcbbuckeye — Sun, 15 May 2011 03:58:01 GMT

I don't usually do much in the way of backtesting, esp with jackpot games. However, I am trying something with P3 on paper. I was reading thru Lotto Logix and came across the Power Trails system. I had read this before but forgotten about it. Anyway, I went back thru April's P3 draws, day and evening and did a workout using the Power Trails system. Came up with 5 combinations to play on paper for May. So far, 1 straight and 3 box hits. In a test like this, backtesting doesn't do any good, but pl... [ More ]

Reply #1

RJOh — Sat, 14 May 2011 18:55:56 GMT

There are lots of ideas about creating systems and back testing them but so far the only back testing I've seen is testing a combination to see if it could have been a winner in past drawings.

If a system pick combinations that have never come up in the past then there is no chance it would have ever won and you don't need to back test to know that. On the other hand if a system only pick combinations that have come up in the past then back testing will show they were winners and we also don... [ More ]

Of Cat Washing and Backtesting

garyo1954 — Sat, 14 May 2011 17:24:48 GMT

When Aaron was three he was told, If you put that cat in water he will scratch you to death! And yet, at some point, he found himself in the bath with a cat under one arm and tub full of water.

When he emerged he looked like he'd been fighting a pride of lions over the last hyena hindquarter. But he was proud. He kept reminding us, ...but it didn't scratch me to DEATH!

Backtesting is the same. Everyone knows, or should know, before you hit GO what the results are going to be. 100% of back... [ More ]