U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > U.S. Forums > New York > New York City
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 07-10-2018, 06:54 AM
 
3,447 posts, read 3,521,842 times
Reputation: 3079

Advertisements

Quote:
Originally Posted by MechaMan View Post
So nobody is exploring the possibility of certain stores or neighborhoods getting more winning tickets than others?

They make it difficult to analyze the winning location trends. There is a reason for that.
There can be one of two reasons for this :

1. You cannot analyze or rationalize random seed. 1 in 5 tickets is a winner with a value just below those 5 tickets combined, 1 in 50,000 is a winner with a value significantly below those 50,000 combined. That big winner is randomized and ends up in a random location. In this case rationalizing anything is a waste of time as the turnout is random.

2. Winning pieces are deliberately placed in certain places because the internal institution is corrupt. Something similar to this happened with the McDonalds Monopoly game over a series of years where internal stakeholders essentially figured out how to steal the top prizes. In this case rationalizing anything is a waste of time as you cannot win since the game is rigged.

Eight charged with defrauding McDonald's for prizes - Chicago Tribune


Any 'advice' the lottery board gives you is in hope to get you to purchase more tickets and play more.
Do research on the Gamblers Fallacy. It might bring you some comfort in all this.
Reply With Quote Quick reply to this message

 
Old 07-10-2018, 02:23 PM
 
1,063 posts, read 324,520 times
Reputation: 1423
Quote:
Originally Posted by WithDisp View Post
There can be one of two reasons for this :

1. You cannot analyze or rationalize random seed. 1 in 5 tickets is a winner with a value just below those 5 tickets combined, 1 in 50,000 is a winner with a value significantly below those 50,000 combined. That big winner is randomized and ends up in a random location. In this case rationalizing anything is a waste of time as the turnout is random.
Thank you and makes sense, this is one of the types of answers I was looking for!
Reply With Quote Quick reply to this message
 
Old 07-10-2018, 03:28 PM
 
8,230 posts, read 8,512,384 times
Reputation: 10197
I don't play the lottery, but a few year ago I was given a few scratch-off cards as a birthday present.

It occurred to me then how easy it would be to totally cheat the public without their knowing. When you pick numbers, they have to have a drawing. But with scratch-off, let's say that the realized today that they had been losing an unusually large amount of money given out I winnings. It would be simple for them to just cut back drastically on the number of winners they print for the rest of the year.
Reply With Quote Quick reply to this message
 
Old 07-11-2018, 08:17 AM
 
Location: crafton pa
980 posts, read 301,594 times
Reputation: 1183
Quote:
Originally Posted by MechaMan View Post
So nobody is exploring the possibility of certain stores or neighborhoods getting more winning tickets than others?

They make it difficult to analyze the winning location trends. There is a reason for that.
That's a much more complicated analysis than you might think. Assuming you have data on where the big winners were purchased, it can be done, but it likely would not show anything definite. Just to give you an example, suppose you theorized that a roulette wheel at a casino comes up with the number 36 more often than the other numbers. (If we ignore 0 and 00 on the wheel, this is mathematically analogous to numbering your neighborhoods 1-36 and theorizing that neighborhood 36 gets more big winners than the other neighborhoods). How would you go about proving it? Well, you could record the results from that wheel. There are problems here, though. Suppose you watch 10 spins and see 36 come up twice. Is that enough to prove your theory, or could that have just happened by random chance? Anyone who's ever gone to a casino and plays roulette would probably tell you that seeing the same number come up twice in 10 spins is not really all that uncommon. Mathematically, for any given number, it's about a 2% chance that you'd see two occurrences of that number in 10 spins.

That seems pretty low, but you have to consider, you haven't specified which neighborhood you think is the one with more winners. Therefore, the relevant probability is the probability of ANY number occurring twice in 10 spins, which is considerably higher, about 75%. Certainly, this provides no proof of anything.


The next step, of course, is to continue to watch the results and see if the frequency of 36 comes out higher than expected (1/36 is the expected value). How much higher does it have to be to become meaningful? Again, that's a tough question, and there always is the possibility that no matter how unlikely it would be, the wheel actually is unbiased and that the result you see is simply the result of random chance. Statistics can provide some guidelines, but it will always be probabilistic guidelines. That is, we can say there's a 0.01% chance of these results having come randomly from an unbiased roulette wheel (or some other number). We can never say "this roulette wheel is definitely biased."


There is a further complication that comes with the actual problem, rather than with the roulette wheel model. The roulette wheel model is simplified because the expected probability of each possible outcome is the same. That is, we have no reason to believe that 1, 2, 3, 36 or any other number should occur more often than any other value. In fact, real roulette wheels are intentionally designed with this in mind; a biased roulette wheel is a potential disaster for a casino (There actually was a real life case in a European casino where some mathematicians detected a bias and made a good bit of money from it). In the "which neighborhoods are there more winners" problem, by contrast we expect the frequencies of winners to be different even if there is no bias in the distribution of tickets.

Consider, let's divide NYC into five neighborhoods, corresponding to the five boroughs and assume for simplicity that all NYC residents are equally likely to buy tickets. Suppose we observe 50 lotteries and find that there were 8 winners in the Bronx, 13 winners in Queens, 11 winners in Manhattan, 14 winners in Brooklyn, and 4 winners in Staten Island. Does this prove that the lottery is biased against Staten Island? Of course not! We would expect fewer winners in Staten Island simply because we expect fewer tickets to be sold there. In fact, based on my assumptions, SI is overrepresented--we should expect only about 3 winners there. Again, statistics can help us answer questions about how different from expected the distribution of winners would have to be before we would think something might be amiss (google the Chi squared distribution for more info), but again, the answer is only probabilistic; statistics can never say that something non-random is definitely happening, only that the probability is low of seeing these data if the distribution really is random.
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply
Please update this thread with any new information or opinions. This open thread is still read by thousands of people, so we encourage all additional points of view.

Quick Reply
Message:



Over $104,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > U.S. Forums > New York > New York City
View detailed profiles of:
Follow City-Data.com founder on our Forum or

All times are GMT -6.

2005-2018, Advameg, Inc.

City-Data.com - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35 - Top