Welcome Guest
You last visited December 4, 2016, 7:21 pm
All times shown are
Eastern Time (GMT-5:00)

# Fooled by Randomness

Topic closed. 297 replies. Last post 6 years ago by jimmy4164.

 Page 10 of 20

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 18, 2010, 4:59 pm - IP Logged

"...how many hits will (future tense) a three digit number repeat over a 33.4 year period..."

OK, maybe I should have said "did" instead of "will."

I thought it was clear from all the other verbage that it was a backtest of Draws made previously.  This point doesn't alter any of the conclusions drawn.

Kentucky
United States
Member #32652
February 14, 2006
7298 Posts
Offline
 Posted: August 18, 2010, 5:27 pm - IP Logged

"...how many hits will (future tense) a three digit number repeat over a 33.4 year period..."

OK, maybe I should have said "did" instead of "will."

I thought it was clear from all the other verbage that it was a backtest of Draws made previously.  This point doesn't alter any of the conclusions drawn.

It was obvious you meant "should" and no it doesn't alter any conclusions.

I finally found an exclusive PA evening data base where I can examine other probabilities over the 33 year drawing history. I can't check the top hits from the last 5400 drawings so I can't say for a fact 25 of them will show a profit.

Kentucky
United States
Member #32652
February 14, 2006
7298 Posts
Offline
 Posted: August 18, 2010, 6:58 pm - IP Logged

Continuing:

"1) DID YOU  PREDICT THESE 25 NUMBERS WITH ONE OF YOUR systems, OR DID YOU MERELY RETRIEVE THEM FROM THE RESULTS, after the fact, LIKE I DID?"

I pretty much answered how systems are created from back testing and acknowledged the fact had the 527 birthday boy played that number from the beginning, he would have probably lost money. Haven't seen all the systems but the majority I have seen are based on past drawings so of course they'll show a profit after the fact.

"2) When you say, "Made Money," JUST HOW MUCH MONEY ARE YOU TALKING ABOUT?"

Making money was just an observation based on the fact 25 combined drawing numbers were drawn 11 or more times and the payoff exceeded the cost of play. If I had a system that could predict which numbers will hit twice or more in the next 500 drawings, I'd be talking about a substantial profit.

"And oh, BTW, check my FIRST post at this site and you will find that I brought up the Nick Perry FIASCO in Pittsburgh then, so you're kind of late bringing it up here.  That fraudulant draw is a tiny blip on a distant horizon!"

And that's why I commented on Perry; the irony of knowing 1 of only 8 numbers will hit and not winning a dime.

"Well STACK47, what have you enlightened us with that we didn't already know? "

In the beginning it sure looked like you were trying to create a system but apparently you changed your mind.

"For anyone who got excited over the 23 hits of the number 308 above, consider this.  If you bought a \$1 ticket on 308 every day for the 30+ years, you would have spent \$11,567.00.  Your 23 hits would have netted you \$11,500(In PA 23*500), for a NET LOSS of \$67.00."

Simple math shows that PA and all state lotteries paying \$500 for \$1 only payoff 49.9% of the true payoff and doubt anyone with a slight knowledge of probability would have thought even one number would average 2 hits per 1000 drawings in almost 12,000 drawings.

Please enlighten me on something I don't know. Do you actually know someone that played the same number every night for 33 years?

United States
Member #59354
March 13, 2008
3962 Posts
Offline
 Posted: August 19, 2010, 2:03 am - IP Logged

RL,

I'm glad you've reduced this problem to it's simplest case - 1:3 .

"Lets say that we have three choices and must select 1.  On average I will select the correct one 1 of every 3 attemps."

Correct.  (That is IF we NOW assume the Lottery has the same set of 3 balls in its hopper, and ONLY those.)

"If I remove one I now have a 50/50 decision to make for 66% chance of being correct."

Wrong!  You still ONLY have a 33% chance to win.  You ONLY have a better chance of picking the ball that will match what the lottery is going to select FROM ITS SET OF 3 -- IF YOU HAVE PRIOR KNOWLEDGE THAT THE ONE YOU REMOVED IS NOT GOING TO BE THE ONE THEY SELECT!!!!!!!!!!!!!!!!!!!!!!!!!!

YOU DON'T HAVE THIS PRIOR KNOWLEDGE!

THIS IS NOT THE MONTY HALL SHOW!!

Monte is not here to assist you by effectively telling you which ball to remove!

If he was, you would be absolutely correct!

Monte KNEW what was behind all 3 doors, REMEMBER?

Now I see your problem!  You are making the mistake that 92% of the population made years ago evaluating the Monte Hall Problem , BUT IN REVERSE!

--Jimmy4164

p.s.  I really hope this a "Eureka Moment" for you!

Jimmy

If you would have read the post you would have seen that I made the

statement that each selection had a 3.3 chance the, "." was a typo.

I have been selecting  3 numbers not to be drawn in MO. Pick-3 game

Each number is positional.

I have been doing this for several drawings and not one draw has matched

a single number that I selected.

Now if I remove one number for each of the 3 positions then I have only

729 possible sets to choose from of the original 1000.

1 in 729 vs 1 in 1000

I am going to keep this up and see how many drawings I can do this without

a miss.

I don't think it wil end in a "eureka moment" at least not for you.  And you are

correct Monty is not here helping.

If you think this has anything to do with the Monty Hall Problem you are way

off.  It is simply a matter of odds.

RL

Working on my Ph.D.  "University of hard Knocks"

I will consider the opinion that my winnings are a product of chance if you are willing to consider

they are not.  Many great discoveries come while searching for something else

Trump / 2016 & 2020

Michigan
United States
Member #22395
September 24, 2005
1583 Posts
Offline
 Posted: August 19, 2010, 12:31 pm - IP Logged

RL,

I'm glad you've reduced this problem to it's simplest case - 1:3 .

"Lets say that we have three choices and must select 1.  On average I will select the correct one 1 of every 3 attemps."

Correct.  (That is IF we NOW assume the Lottery has the same set of 3 balls in its hopper, and ONLY those.)

"If I remove one I now have a 50/50 decision to make for 66% chance of being correct."

Wrong!  You still ONLY have a 33% chance to win.  You ONLY have a better chance of picking the ball that will match what the lottery is going to select FROM ITS SET OF 3 -- IF YOU HAVE PRIOR KNOWLEDGE THAT THE ONE YOU REMOVED IS NOT GOING TO BE THE ONE THEY SELECT!!!!!!!!!!!!!!!!!!!!!!!!!!

YOU DON'T HAVE THIS PRIOR KNOWLEDGE!

THIS IS NOT THE MONTY HALL SHOW!!

Monte is not here to assist you by effectively telling you which ball to remove!

If he was, you would be absolutely correct!

Monte KNEW what was behind all 3 doors, REMEMBER?

Now I see your problem!  You are making the mistake that 92% of the population made years ago evaluating the Monte Hall Problem , BUT IN REVERSE!

--Jimmy4164

p.s.  I really hope this a "Eureka Moment" for you!

Just for fun...

This link (if Todd leaves it here) is a simulation based on Monty Hall NOT knowing what is behind the doors.  There is also a link at the site with the original Monty Hall problem where he does know what is behind the doors.

Let's Make a Deal: Screen 2:
http://math.ucsd.edu/~crypto/cgi-bin/MontyDoesNotKnow/monty2?0+15030

Kentucky
United States
Member #32652
February 14, 2006
7298 Posts
Offline
 Posted: August 19, 2010, 1:49 pm - IP Logged

Jimmy

If you would have read the post you would have seen that I made the

statement that each selection had a 3.3 chance the, "." was a typo.

I have been selecting  3 numbers not to be drawn in MO. Pick-3 game

Each number is positional.

I have been doing this for several drawings and not one draw has matched

a single number that I selected.

Now if I remove one number for each of the 3 positions then I have only

729 possible sets to choose from of the original 1000.

1 in 729 vs 1 in 1000

I am going to keep this up and see how many drawings I can do this without

a miss.

I don't think it wil end in a "eureka moment" at least not for you.  And you are

correct Monty is not here helping.

If you think this has anything to do with the Monty Hall Problem you are way

off.  It is simply a matter of odds.

RL

"Now if I remove one number for each of the 3 positions then I have only  729 possible sets to choose from of the original 1000. 1 in 729 vs 1 in 1000."

A few pages back I mentioned other probabilities and the cause and effect each digit has on the others, but was told it was irrelevant because it had "nothing to do with Jim's statistics. However there are three individual drawings and it's impossible for all three digits to repeat if any one of them doesn't. With Jim's statistics, each digit position does have an effect on the others. The odds against repeating are 9 to 1 in each digit position, 9 times 9 times 9 = 729 to 1.

My program only allows me analyze the last 7500 PA evening pick-3 drawings (over 20 years) but it's certainly enough data to analyze each digit position. The first digit repeated 788 times, the second 754, and the third 760. From this we know there were 754 chances out of 7500 or a little bit over 10% for all three digits to repeat. Two digits repeated on the same drawing in any position 239 times, 31.8% of the chances but in only 3.2% of all the drawings.

For practical playing purposes, I can't see anyone playing the three digit number from the previous drawing knowing they should have only 3 or 4 chances in the next 100 drawings. How many times should a three digit number repeat in 20 years of drawings? (7500 divided by 729 = 10.28) How many times did it happen in the last 7500 drawings? (10 times)

It's an obvious losing if not very boring betting proposition for 7500 drawings and it only gets worse playing it in all 11,500 drawings. It should have happened 15 or 16 times but only happened 14.

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 19, 2010, 7:22 pm - IP Logged

"Now if I remove one number for each of the 3 positions then I have only  729 possible sets to choose from of the original 1000. 1 in 729 vs 1 in 1000."

A few pages back I mentioned other probabilities and the cause and effect each digit has on the others, but was told it was irrelevant because it had "nothing to do with Jim's statistics. However there are three individual drawings and it's impossible for all three digits to repeat if any one of them doesn't. With Jim's statistics, each digit position does have an effect on the others. The odds against repeating are 9 to 1 in each digit position, 9 times 9 times 9 = 729 to 1.

My program only allows me analyze the last 7500 PA evening pick-3 drawings (over 20 years) but it's certainly enough data to analyze each digit position. The first digit repeated 788 times, the second 754, and the third 760. From this we know there were 754 chances out of 7500 or a little bit over 10% for all three digits to repeat. Two digits repeated on the same drawing in any position 239 times, 31.8% of the chances but in only 3.2% of all the drawings.

For practical playing purposes, I can't see anyone playing the three digit number from the previous drawing knowing they should have only 3 or 4 chances in the next 100 drawings. How many times should a three digit number repeat in 20 years of drawings? (7500 divided by 729 = 10.28) How many times did it happen in the last 7500 drawings? (10 times)

It's an obvious losing if not very boring betting proposition for 7500 drawings and it only gets worse playing it in all 11,500 drawings. It should have happened 15 or 16 times but only happened 14.

Stack47 AND RL-RandomLogic,

I WAS NOT proposing that anyone play any particular number - yesterday's winner, or whatever.  What I had hoped would happen before now is that at least one person would have observed that it really doesn't matter how you choose your Straight Pick on any given night.  You can go to the PA website and learn how many times 1-3-5, 2-4-6, 3-5-7,,,,,,5-7-9 hit, and I predict you will get the same results we got for the other trials:  12 with a 3.3 S.D.  Then try other [what some regard as non-random] patterns of your choice.  This is why QPs produce winners in relation to their percentage of participation!

(The PA site is quirky - you must FORCE it to deal with Evening - it LIKES Midday best)

Why can't you shake off the Gambler's Fallacy?  In the past, I couldn't understand how anyone could SLOWLY, with their critical thinking cap on, read an analysis of the Fallacy, like the one at Wikipedia, and come away still believing in it.  I was wrong.  That's why I looked for psychological explanations.  I sent you links to what I found at the beginning of my search, but you've made no comment on them.  RL claims the Gambler's Fallacy is not really a fallacy.  Might I point out that Wikipedia has an arbitration and editing process that anyone can use to challenge an entry in their encyclopedia.  I KNOW what would happen if challenged this principal, and it wouldn't be good for your ego.  Why not print out the wiki on the fallacy and the psychological article I linked you to and find a cool quiet place to digest them.

--Jimmy4164

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 19, 2010, 7:59 pm - IP Logged

Stack47 AND RL-RandomLogic,

I WAS NOT proposing that anyone play any particular number - yesterday's winner, or whatever.  What I had hoped would happen before now is that at least one person would have observed that it really doesn't matter how you choose your Straight Pick on any given night.  You can go to the PA website and learn how many times 1-3-5, 2-4-6, 3-5-7,,,,,,5-7-9 hit, and I predict you will get the same results we got for the other trials:  12 with a 3.3 S.D.  Then try other [what some regard as non-random] patterns of your choice.  This is why QPs produce winners in relation to their percentage of participation!

(The PA site is quirky - you must FORCE it to deal with Evening - it LIKES Midday best)

Why can't you shake off the Gambler's Fallacy?  In the past, I couldn't understand how anyone could SLOWLY, with their critical thinking cap on, read an analysis of the Fallacy, like the one at Wikipedia, and come away still believing in it.  I was wrong.  That's why I looked for psychological explanations.  I sent you links to what I found at the beginning of my search, but you've made no comment on them.  RL claims the Gambler's Fallacy is not really a fallacy.  Might I point out that Wikipedia has an arbitration and editing process that anyone can use to challenge an entry in their encyclopedia.  I KNOW what would happen if challenged this principal, and it wouldn't be good for your ego.  Why not print out the wiki on the fallacy and the psychological article I linked you to and find a cool quiet place to digest them.

--Jimmy4164

Let's say the last four nights of some state lottery's Pick-3 produced the following sequence:

2010/08/17    3 - 7 - 2

2010/08/18    4 - 1 - 3

2010/08/19    4 - 1 - 3

Now, the probability of 4-1-3 hitting 2 days in succession (Calculated on 8/17) was:

( 1 / 1,000 ) * ( 1 / 1,000 )   (One over one thousand - squared)

Which is ( 1 : 1,000,000 )    (One in a Million)

Now that this has happened, you have to make a choice.

Is 4-1-3 a reasonable choice for tommorows drawing?

MANY would argue NO, because the odds of getting 4-1-3 three days in a row would be:

( 1 / 1,000,000 ) * ( 1 / 1000 )   (The odds of 2 occurences TIMES the odds of one more.)

Which is 1 : 1,000,000,000   (One in a Billion)

But they would be commiting the Gambler's Fallacy,

AND THEY WOULD BE WRONG!

TOMORROW, the odds of 2 consecutive drawings of 4-1-3 on August 18th and 19th will be ONE ( 1.0 ) !!!

IT'S A CERTAINTY (1.0) - BECAUSE THEY HAVE ALREADY BEEN DRAWN!!!!!

The odds of 4-1-3 being drawn TOMMORROW is  ( 1 : 1,000 )     :-(

--Jimmy4164

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 19, 2010, 8:26 pm - IP Logged

Is random the same as random or is there something as more random and or less random, if the drawings are "random" why have pre-draws and have changes of machines and ball sets, Does doing those things make draws more random? Does that means that the drawings are not random to start with and that they need some "help" so they can be made random?

The spins inside the drum are not enough to make the drawings random? If that is not enough then nothing would be enough, perhaps if that is true then there is no such thing as random.

Are "Random Patterns" a valid notion? If there is something as "Random Patterns" then it might mean that "Random" has patterns, if so, then can there be really such as things are random? Isn't "random" a lack of patterns?

And if there are patterns and if there is no random, there perhaps people who use stats of patterns can get some winning advantage on the games.

-----------

Never mind any of that, sometimes, often or all of the time it might be best to leave people or some alone with their own opinions.

These things can have no end to them and drag on for ever.

Let people be can be both ways:

2: Don't tell.

= Don't argue.

Maybe that doesn't = Everybody happy or it does, but Who can have it all? Maybe some can?

So long as Be and Let Be doesn't hurt in some way one side or the other that might at least mean peace.

------------

But Who am I to say those things?, If it didn't bother me, I would not be posting this.

The routine changing of ball sets and machines is to ensure that if there are some tiny differences in ball weight, ball diameter, or machine interior sufaces, they will not propagate a bias over time.  The pretests are simply to ensure the equipment is working properly.  Don't you think it would be [more than] a little embarrassing if the LIVE TV production crew rolled the cameras and a short in a ball machine flipped a circuit breaker, or a ball got stuck inside a machine, or a computer started spewing sparks and/or smoke!

As far as the effects on the randomness of the results:  If the ball equipment or the RNGs have been vetted as truly random, there should be NO effect.  NOW, if, while the cameras are rolling, ball number 3 pops out of machine number 1, and an official throws up his/her hands and calls foul, puts the ball back in the machine, and recycles it - THEN THERE MIGHT BE A PROBLEM!!

Michigan
United States
Member #22395
September 24, 2005
1583 Posts
Offline
 Posted: August 19, 2010, 9:26 pm - IP Logged

@ jimmy4164

Not sure I fully understand various distinctions.

...it really doesn't matter how you choose your Straight Pick on any given night.

The choices are, choose it by some random method.  Are all random methods going to come up with 12 plus/minus 3.3?  I mean like a quick pick, different computerized randoms, random numbers from this website?   In what time frame?  1 month, 1 year, 11 years (like your PA data)?

RANDOM.ORG - True Random Number Service:
http://www.random.org/

Choose it by some self-pick method not necessarily mathematical based.  This would be by license plates, birthdays, etc;

Choose it by some mystical method.  This would include dreams, fortune tellers or some such.

Choose it based on "patterns" method.  This would include hot, cold, due and any other type of noticeable pattern.

Are you saying ALL of the above will only provide 12 plus/minus 3.3?

Then what do you call the person that uses QPs and gets totally destroyed and does not see one single straight pick 3?  Or the opposite, a QP player that gets way more than his/her share?

And what do you call the opposite where someone picks their own numbers by some strategic method and gets 20 straight pick 3s?  And continues to do so?

Just unlucky and lucky?  Nothing more?  Nothing for the skill of developing something that can outperform quick picks and get 20 winners?  Where is the credit in the gambler's fallacy for the winners?  Let's extend it to all professional gamblers that have for years, defied the odds and won millions?  How does the gambler's fallacy allow that to happen?

Gambler's fallacy is not a law like gravity - it has no way to force a person to win within it's parameters.  Or does it?

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 20, 2010, 12:38 am - IP Logged

truecritic,

"The choices are, choose it by some random method.  Are all random methods going to come up with 12 plus/minus 3.3?  I mean like a quick pick, different computerized randoms, random numbers from this website?   In what time frame?  1 month, 1 year, 11 years (like your PA data)?"

I've been working with 33+ years of the PA Daily (Evening) [000-999].

With this data, the hit rate for arbitrary patterns (like yesterday's winner) has been close to the 12 plus/minus a 3.3 std dev, which was calculated by looking at all 1000 numbers over that period.  (Which, BTW, implies that QPs will do the same!   ) (Later!)

Remember, 3.3 is ONE(1) Standard Deviation.  We found one number (308) that hit 23 times, which is one hit outside 3 Standard Deviations.

When you say "all random methods" you make it difficult to answer your question.  I would be very surprised if QPs or SPs or Random.org were not random enough to suffice for statistical predictions on the likes of lottery results.  Time frame is more important.  If you're tossing dice that have been certified "fair," you wouldn't take them back to the seller if you happened to throw [a total of] 7 three times in a row, but if you threw seven 400 out of 500 tosses, you might take them back.  There are mathematical ways to calculate what are called Confidence Intervals which help you decide how many trials you need to be sure you have enough for "statistical significance."  I'm just guessing, but I'm pretty sure my 11,600 (33+ years) worth of 3 digit numbers is enough for what we're doing here.  If you put some of these terms into Google, you can read all about it!

I'm really glad you pointed to random.org; I wasn't aware of it.  It really looks like the numbers from there will be about as "Truly Random" as you can get.  Watch for me to suggest this site soon when I post a simulation of a Lottery I'm designing and try to get people to play it.  This [hopefully] will solve the problem I was anticipating of people rejecting the results by claiming the inputs weren't random.  This will make it more interesting.  Thanks!

The other methods are not as reliable for various reasons.  For example, what do you do about the people who keep having the same dream over and over?

Some here have expressed a mistrust of Wikipedia.  Please do me a favor and plug "Gambler's Fallacy" into your Google search box and take a look at how other sites deal with it.  I prefer Wikipedia for these kinds of issues because of the strict vetting the articles get, but there are other places to get information.  The "Gambler's Fallacy" is at the heart of most of the debates I've followed here at this site over the last month.  The link below is to my attempt to wrench 2 other posters here out of the clutches of the fallacy by use of an example.  But check out the article suggestions at Google first.

I really would like to get some feedback on what you find, and what you think of what you find!

--Jimmy4164

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 20, 2010, 1:54 am - IP Logged

Continuing:

"1) DID YOU  PREDICT THESE 25 NUMBERS WITH ONE OF YOUR systems, OR DID YOU MERELY RETRIEVE THEM FROM THE RESULTS, after the fact, LIKE I DID?"

I pretty much answered how systems are created from back testing and acknowledged the fact had the 527 birthday boy played that number from the beginning, he would have probably lost money. Haven't seen all the systems but the majority I have seen are based on past drawings so of course they'll show a profit after the fact.

"2) When you say, "Made Money," JUST HOW MUCH MONEY ARE YOU TALKING ABOUT?"

Making money was just an observation based on the fact 25 combined drawing numbers were drawn 11 or more times and the payoff exceeded the cost of play. If I had a system that could predict which numbers will hit twice or more in the next 500 drawings, I'd be talking about a substantial profit.

"And oh, BTW, check my FIRST post at this site and you will find that I brought up the Nick Perry FIASCO in Pittsburgh then, so you're kind of late bringing it up here.  That fraudulant draw is a tiny blip on a distant horizon!"

And that's why I commented on Perry; the irony of knowing 1 of only 8 numbers will hit and not winning a dime.

"Well STACK47, what have you enlightened us with that we didn't already know? "

In the beginning it sure looked like you were trying to create a system but apparently you changed your mind.

"For anyone who got excited over the 23 hits of the number 308 above, consider this.  If you bought a \$1 ticket on 308 every day for the 30+ years, you would have spent \$11,567.00.  Your 23 hits would have netted you \$11,500(In PA 23*500), for a NET LOSS of \$67.00."

Simple math shows that PA and all state lotteries paying \$500 for \$1 only payoff 49.9% of the true payoff and doubt anyone with a slight knowledge of probability would have thought even one number would average 2 hits per 1000 drawings in almost 12,000 drawings.

Please enlighten me on something I don't know. Do you actually know someone that played the same number every night for 33 years?

Stack47,

"I pretty much answered how systems are created from back testing and acknowledged the fact had the 527 birthday boy played that number from the beginning, he would have probably lost money. Haven't seen all the systems but the majority I have seen are based on past drawings so of course they'll show a profit after the fact."

This is a very common approach in the stock trading "System" business too.  They optimize the parameters of their system on a particular time frame and then report the results in a way that implies it will perform that well on future inputs that the system has never seen.  Of course, it seldom does.  I'm not surprised the same approach is used in the lottery prediction business.

"And that's why I commented on Perry; the irony of knowing 1 of only 8 numbers will hit and not winning a dime. "

I missed that one!  Gotcha now!

"In the beginning it sure looked like you were trying to create a system but apparently you changed your mind."

I had no intention of creating a system.  It never occured to me since I don't think there is any way to build an edge to get you any more than the 50% the lotteries are paying back.  I am, however, now designing a scaled down Lotto Game, one that can be played to demonstrate transparently the probability principals involved, using a much smaller "Matrix."

This was my first post to this site.

It never got very far, but it was my reason for seeking a site like this.  Because of my background in math and computer science I am skeptical of computerized Draws, and can't think of any way I can be convinced otherwise.  For this reason, I AM sympathetic to the motives of many of the system players, the ones who think they can exploit possible fraud by looking for patterns.  But I think their efforts are doomed to failure.  As a programmer, I know how subtle and brief a ruse could be, and trying to second guess it would require the kind of skills and computer power of the NSA code cracking unit!  In many ways, the effort is even more difficult.  The NSA is usually dealing with a document or a bit stream that they can grab and try to decipher.  From what I've seen discussed here, lottery system players are playing mix and match matrix manipulation, often with no rationale or rhyme or reason.  If there is larceny in the hearts of Lottery Commission programmers, I don't think they'll be caught with these haphazard methods.  That said, I don't want them to stop scrutinizing the results.  I just want them to understand better the probabilities so they can build a better "mouse trap," one that might some day catch a mouse!

My last paragraph is why I think we should lobby for LIVE TV coverage of public drawings made with mechanical machines.  A 50% return is a joke as it is, so why should we pay more for our chance at a Jackpot?

This is one of the main reasons why a site like Lottery Post is a valuable contribution to the whole lottery process.  Through the collective efforts of all the members, a lot of scrutiny is put on the Lottery Commissions, making it less likely they will get greedy!  The other reasons are social.  The management here keeps it clean, but doesn't seem to prohibit people from meandering off on philosophical or other tangents.  This is good.  People who like dealing with numbers typically are more thoughtful than most, and appreciate an outlet for their thoughts.

"Please enlighten me on something I don't know. Do you actually know someone that played the same number every night for 33 years?"

No, I don't.  But my simulations tell me that if they had, they would probably have done just as well as a QP player!

I do have a cousin who was the member of a small group of computer people who won a large jackpot years ago.  He's been living on it ever since.  They generated their own weekly QPs (not a system) using their big old mainframe computer!

--Jimmy4164

Michigan
United States
Member #22395
September 24, 2005
1583 Posts
Offline
 Posted: August 20, 2010, 5:29 am - IP Logged

truecritic,

"The choices are, choose it by some random method.  Are all random methods going to come up with 12 plus/minus 3.3?  I mean like a quick pick, different computerized randoms, random numbers from this website?   In what time frame?  1 month, 1 year, 11 years (like your PA data)?"

I've been working with 33+ years of the PA Daily (Evening) [000-999].

With this data, the hit rate for arbitrary patterns (like yesterday's winner) has been close to the 12 plus/minus a 3.3 std dev, which was calculated by looking at all 1000 numbers over that period.  (Which, BTW, implies that QPs will do the same!   ) (Later!)

Remember, 3.3 is ONE(1) Standard Deviation.  We found one number (308) that hit 23 times, which is one hit outside 3 Standard Deviations.

When you say "all random methods" you make it difficult to answer your question.  I would be very surprised if QPs or SPs or Random.org were not random enough to suffice for statistical predictions on the likes of lottery results.  Time frame is more important.  If you're tossing dice that have been certified "fair," you wouldn't take them back to the seller if you happened to throw [a total of] 7 three times in a row, but if you threw seven 400 out of 500 tosses, you might take them back.  There are mathematical ways to calculate what are called Confidence Intervals which help you decide how many trials you need to be sure you have enough for "statistical significance."  I'm just guessing, but I'm pretty sure my 11,600 (33+ years) worth of 3 digit numbers is enough for what we're doing here.  If you put some of these terms into Google, you can read all about it!

I'm really glad you pointed to random.org; I wasn't aware of it.  It really looks like the numbers from there will be about as "Truly Random" as you can get.  Watch for me to suggest this site soon when I post a simulation of a Lottery I'm designing and try to get people to play it.  This [hopefully] will solve the problem I was anticipating of people rejecting the results by claiming the inputs weren't random.  This will make it more interesting.  Thanks!

The other methods are not as reliable for various reasons.  For example, what do you do about the people who keep having the same dream over and over?

Some here have expressed a mistrust of Wikipedia.  Please do me a favor and plug "Gambler's Fallacy" into your Google search box and take a look at how other sites deal with it.  I prefer Wikipedia for these kinds of issues because of the strict vetting the articles get, but there are other places to get information.  The "Gambler's Fallacy" is at the heart of most of the debates I've followed here at this site over the last month.  The link below is to my attempt to wrench 2 other posters here out of the clutches of the fallacy by use of an example.  But check out the article suggestions at Google first.

I really would like to get some feedback on what you find, and what you think of what you find!

--Jimmy4164

I got mixed up and said 11 years instead of 33 years and 11,000+ draws.  So I should have asked:  In what time frame?  1 month, 1 year, 33 years (like your PA data)?

Sticking to Pick 3, not jackpot games.

If you can only call it luck, this is where we disagree.  I say it is based on skill to be able to pick 3 numbers and beat the odds - based on a system or method that a person uses.

I read Gambler's Fallacy there when you posted it originally (back when you joined LP)...that is why I had these questions about it.

Let's extend it to all professional gamblers that have for years, defied the odds and won millions?  How does the gambler's fallacy allow that to happen?

Gambler's fallacy is not a law like gravity - it has no way to force a person to win within it's parameters.  Or does it?

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 20, 2010, 10:30 pm - IP Logged

I got mixed up and said 11 years instead of 33 years and 11,000+ draws.  So I should have asked:  In what time frame?  1 month, 1 year, 33 years (like your PA data)?

Sticking to Pick 3, not jackpot games.

If you can only call it luck, this is where we disagree.  I say it is based on skill to be able to pick 3 numbers and beat the odds - based on a system or method that a person uses.

I read Gambler's Fallacy there when you posted it originally (back when you joined LP)...that is why I had these questions about it.

Let's extend it to all professional gamblers that have for years, defied the odds and won millions?  How does the gambler's fallacy allow that to happen?

Gambler's fallacy is not a law like gravity - it has no way to force a person to win within it's parameters.  Or does it?

truecritic,

"I say it is based on skill to be able to pick 3 numbers and beat the odds - based on a system or method that a person uses."

Here is the request I made of you in the post you replied to above:

-----Jimmy4164 Quote-----

"Some here have expressed a mistrust of Wikipedia.  Please do me a favor and plug 'Gambler's Fallacy' into your Google search box and take a look at how other sites deal with it.  I prefer Wikipedia for these kinds of issues because of the strict vetting the articles get, but there are other places to get information.  The 'Gambler's Fallacy' is at the heart of most of the debates I've followed here at this site over the last month.

.....................

I really would like to get some feedback on what you find, and what you think of what you find!

-------------------------

Based on your current remarks, I think the odds are pretty good that you have NOT YET done me the favor!

Here's a link to the Psychology behind it...

http://en.wikipedia.org/wiki/Gambler%27s_fallacy#Psychology_behind_the_fallacy

See especially The Illusion of Control link.

PULEEEEEZZ???

Once you UNDERSTAND the Gambler's Fallacy, and it sinks in, you will never make your remark [in blue] at the top again!

--Jimmy4164

P.S.

You also asked, "Let's extend it to all professional gamblers that have for years, defied the odds and won millions?  How does the gambler's fallacy allow that to happen?"

Professional gamblers, like stock traders, horse race handicappers, sports betters, blackjack card counters, etc., are NOT GUILTY of believing the Gambler's Fallacy!  Why?  Because the historical events and statistics that they rely on to predict the short term future have been shown(PROVED) to have predictive abilities, based on cause and effect relationships between them and subsequent events.  When a Blackjack player updates his/her counters when cards are removed from play, he/she has a slightly better edge than before, and their edge continues to increase UNTIL THE CARDS ARE RESHUFFLED!  In lotteries, ALL  OF THE "CARDS" ARE REPLACED AND RESHUFFLED AFTER EVERY DRAW!!!

How can I make it any clearer than this?

United States
Member #93947
July 10, 2010
2180 Posts
Offline
 Posted: August 21, 2010, 1:17 am - IP Logged

Just for fun...

This link (if Todd leaves it here) is a simulation based on Monty Hall NOT knowing what is behind the doors.  There is also a link at the site with the original Monty Hall problem where he does know what is behind the doors.

Let's Make a Deal: Screen 2:
http://math.ucsd.edu/~crypto/cgi-bin/MontyDoesNotKnow/monty2?0+15030

truecritic,

The UCSD link you pasted doesn't work for me when I click on it in IE while viewing this page, but I was able to cut/paste it in and access it.  Pretty slick little demo!  The problem with alternative rules to the game is that it gives those who still can't understand [or can't admit they were wrong] why the player should switch in the original stated problem, a way to muddy the waters and deny it's the correct answer.  I know; it caused me to lose a friend over it.

The one I like best is this YouTube animation:

This was my first exposure to the problem:

http://theweek.com/article/index/43224/the_last_word_what_are_the_odds

RL, It appears we were not on the same page in our last couple exchanges.  You kept talking about selecting things for discard that I thought you were going to bet on!  So, when you said that if you selected 3 balls (out of 10) it would be a fair claim that you would win 67% (why not 70?) of the time, I ran the numbers assuming you were betting 3 of 10, and the Lottery was picking 1 of 10.  So, it's only fair that I rework those numbers to correspond to your example of betting on 7, 10,000 times.

This is the original [mistaken by me] betting plan for your System simulation.

Cost Of Tickets                 \$30,000   (10,000 Draws X 3 Tickets per Draw)

Winnings                          \$20,394   (3399 Wins X \$6 per Win)

NET LOSS                          \$9,606     (Cost - Winnings)

Here is the correction, assuming you bet on the 7 left after discarding 3.

Cost Of Tickets                 \$70,000   (10,000 Draws X 7 Tickets per Draw)

Winnings                          \$47,586   (7931 Wins X \$6 per Win)  [Based on your 3399 run.]

NET LOSS                         \$22,414   (Cost - Winnings)

It seems that no matter what you do you're going to lose 40% of your money in MO!  Of course, I know that ignores that your system is supposed to discard digits that the Lottery Commission or ball machines don't like.  Sorry!

Maybe somebody will see why I made an analogy between your probability claims here and the REVERSE error in the Monte Hall problem.  I agree, it's not a very good analogy!

...truecritic - I hope you'll soon be done doing that "favor" for me!

--Jimmy4164

 Page 10 of 20