Page 1 of 2

Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 1:55 am
by Eetrab
Recently, a friend proposed a gambling scenario and asked if I was interested in participating. The description is as follows:

Entering the game costs $1. Once entered, your chance of success is based on a spinner*. The is divided into four equal sections, each with (an equal) chance of landing on. The four outcomes are either losing $3, $2 or gaining $3 or $5, depending on what segment the spinner landed on.
If you caught this, congratulations. I am interested to see how different people's intuition or presumed knowledge of statistics differs from the actual outcomes.
My question is, would you participate in this game? Do you think you have a better chance of gaining money over a number of turns? I know what I would do (PM me for my answer and why).

Please, base your vote on what you think, not on what others think. This is not a trick question (there is no wrong or write answer).

*For those ultra-geeks: the spinner is completely unbiased.

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 2:56 am
by Tee
The question requires clarification. When you say the outcomes are "losing" or "gaining", does this take into account the entry fee?

So if I buy into the game for $1 and the spinner lands on the best outcome, do I receive a total of $5, or do I receive my entry fee back plus $5 for a total of $6?

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 4:30 am
by mitchie151
Mathematically speaking, if we assume that you pay 1$ to spin each time, you have an equal chance of losing a total of 3 or 4 dollars, or earning 3 or 4. Assuming you play a huge amount of times, your profit/loss would still be 0. Thus, I would not play!

EDIT: misread numbers, you would be in the negative.

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 6:13 am
by cgfirecoral
Assuming (this wasn't clear), that you pay $1 per spin, you'll lose money on the average. Basically, you should be down $1 every four spins. It's a losing proposition. But then again, so is Vegas, and people still go there.

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 6:38 am
by faldette
mitchie151 wrote:Mathematically speaking, if we assume that you pay 1$ to spin each time, you have an equal chance of losing a total of 3 or 4 dollars, or earning 3 or 4. Assuming you play a huge amount of times, your profit/loss would still be 0. Thus, I would not play!

EDIT: misread numbers, you would be in the negative.
Yeah, the loosing spins would net losses of 3 or 4, while the winnings would net 2 or 4. Average is definitely negative, so I would recommend not playing. <.<

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 8:18 am
by OZinky
Eetrab wrote: If you caught this, congratulations. I am interested to see how different people's intuition or presumed knowledge of statistics differs from the actual outcomes.
I win! ;D

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 12:54 pm
by Eetrab
TeeJayDub wrote:The question requires clarification. When you say the outcomes are "losing" or "gaining", does this take into account the entry fee?

So if I buy into the game for $1 and the spinner lands on the best outcome, do I receive a total of $5, or do I receive my entry fee back plus $5 for a total of $6?
Eh, I was tired. When you gain, you receive $X, minus the fee. When you lose, you lose $X and and the fee that you played. Us that clear?

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 1:00 pm
by dlgn
25% chance of losing $3 in addition to the entry fee (-$4*1/4=-$1), 25% chance of losing $2 in addition to the entry fee (-$3*1/4=-$.75), 25% chance of gaining $3 minus the entry fee ($2*1/4=$.50), and 25% chance of gaining $5 minus the entry fee ($4*1/4=$1), totaling -$.25. Therefore, it is not profitable to play.

~dlgn

p.s. *right

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 1:29 pm
by Lucozadia
I understand this is a thought experiment and all but I've never come across any from of gambling whereby some outcomes can cause you to lose more than your initial stake and if such a game did exist then I think the site would discover that most players only had $1 in their account ;)

But yeah as pointed out, you'll average a 75% return on this game long term which is far worse than a roulette wheel for instance, so no, I wouldn't play it even if I felt like gambling.

Re: Statistical thought experiment (of sorts)

Posted: Tue Jul 30, 2013 4:59 pm
by bloodthirsty958
I would be the person hosting the game. The house always wins!