Statistics Question

[ChrisB2]

Obvious answer is obvious.

Take the cash you would have wasted on stupid slot machines and get yourself some hookers and blow.

I would "like" that if it said "get yourself some good tequila and put the rest in the bank."

 

[line6man]

A nickle slot machine averages a 90% payout (you'll average a $90 win for every $100 you spend). Would your payout average half, double or stay the same if you spent the same amount of money playing two slot machines at the same time?

Take this with a grain of salt, because I'm not very good with probability theory, but this is what I think to be correct.

Let X1 and X2 be two independent random variables, such that E(X1)=E(X2)=0.9 (Where E(X) denotes the expectation.) By a well known theorem from probability, we know that the expectation of a product of independent random variables is equal to the product of the expectations of the independent random variables. Hence, E(X1*X2)=E(X1)*E(X2)=(0.9)(0.9)=0.81.

So your expected payout is going to be $81.

Anyone want to correct me, if I'm wrong?

 

[48thStreetCustom]

Ha ha ha. Just what do you need to be convinced?

Oh yeah two women playing multiple slots for hours does sort of negate the science.

Were they wearing their lucky socks?

I'm convinced. How many times do I need to say it?

 

[ShredderMaximus]

I would "like" that if it said "get yourself some good tequila and put the rest in the bank."

What about tequila and a subscription to Brazzers?

 

[Downunderwonder]

I'm convinced. How many times do I need to say it?

Once more with conviction.

 

[Byan]

Everything depends on the machines themselves

 

[12BitSlab]

Slot machines do NOT use random numbers. They use pseudo random numbers provided by a PRNG (pseudo random number generator). Such algorithms are part and parcel of math. Without going into the hairy details of the polynomials used for PRNG's, the pattern is predictable since they are built on linear congruential equations (LCG's). Every LCG has a "period" in which the pattern will repeat. The LCG's used in the 1960's and 70's had periods that could be as low as 30,000. BTW, DOD war game simulations used these weak LCG's during most of the Cold War.

Modern PRNG's have periods that are measured in Octillions given a "random" seed. What is the random seed used on a modern slot machine? The time in 10 Billionths of a second between games MOD an incrementing 64 bit "tick value". And, BTW, each wheels gets it own PRNG. Every game (i.e., one pull of the arm) is 100% independent of the previous game unless you can time your pulls down to less than 0.0000000001 of a second. Good luck with that.

So, the bottom line is that slot machines ARE predictable, but you would have to spend Billions of dollars on shared computing resources in order to reliably win Thousands of dollars. You are much better off putting your money into a high quality mutual fund.

 

[Downunderwonder]

Take this with a grain of salt, because I'm not very good with probability theory, but this is what I think to be correct.

Let X1 and X2 be two independent random variables, such that E(X1)=E(X2)=0.9 (Where E(X) denotes the expectation.) By a well known theorem from probability, we know that the expectation of a product of independent random variables is equal to the product of the expectations of the independent random variables. Hence, E(X1*X2)=E(X1)*E(X2)=(0.9)(0.9)=0.81.

So your expected payout is going to be $81.

Anyone want to correct me, if I'm wrong?

your maths only applies to a simple series of two trials, so basically barking up the wrong tree in the wrong town as far as the OP question goes.

 

[hbarcat]

See, I would think the percentage would go down. Think of it this way. If I went out on six dates with one girl, my odds of getting lucky are like 90%. But if I went out on three dates each with two different girls, my odds of getting lucky are only about 50%. Putting in the time with one slot machine must increase the odds.

The failure of this argument is that the slot machines you're playing with all have the same percentage of payout so that any period of time playing them gets you the $90 back for every $100 spent. If you play 5 different machines, they all have that same payout chance. Even if you play on 50 machines at once you're still getting the same payout chance.

The only way the dates with different girls analogy would be applicable to slot machines is if you knew ahead of time that every girl you went on a date with would produce the same likelihood of a "pay out" as any other girl. This is clearly not the case - or else why bother dating other girls - just pick one and stick with her and be happy with giving her 10% of your cash on every date.

Of course the reason you date different people is to find the one who is right for you. That would be great if that was how a casino worked. You could just try each and every machine to see which ones paid more to the house until you found the special machine that for some reason paid back more to the gambler than it took in. Then you could quit your job and play that machine all day and make a fortune while the casino owners cried and cried as they went steadily broke.

Yeah, that never happens. . . . .