Morgan Stanley Traders Lost $390 Million in One Day in August

[Equalizer]

Quote from Lawrence Chan:

I am trying my best not to point fingers here because I think someone mixed up the usage of statistics ...

1. If there is one and only one event of choosing a cup and then an offer to switch, then it is all about luck and forget about probabilities.

i.e. Go tell the winner of a lottery ticket that he is an idiot wasting $ on something with such low probability.
He can tell you that every single $ he won is real.

2. If you are betting on a series of combo events stated in #1, then do the switch because it will pay off!

i.e. if you do not believe me, write a small program to conduct the test over, say, 1000 cases of random selections.

Trading is about consistencies in exploiting the edge you have. That is probability at work.

Yes, what you are talking about in your 'edge' comment is expectation. Lets not confuse expectation with probability.

If you can't see why probabilistically it always makes sense to switch - strictly under the MontyHall scenario, then draw a phreakin' Venn diagram and consider the event space.

Yes, in a once off event, it will be about luck, there are not enough events to let the Law of Large numbers do its bit with regards to expectation. But who cares. Even if you had 100 cups with 98 cups being winners and you had to choose only one, you can still choose the cup without the ball. So what? Are the odds in your favour? They sure are. Same with the monty hall problem, if you remain with the original choice your probability of winning is 1/3, if you switch it is 2/3.
Now concentrate on the next sentence: Regardless of the eventual outcome.

For goodness sakes, its not Girsanov's theorem!

 

[Equalizer]

 

[bidask]

oops i read the question wrong. you are not choosing again from the beginning.

Quote from bidask:

the answer to this Monty Hall's paradox is absolutely correct but it's another one of those academic exercises that rely on an academic assumption. the assumption is that you get to travel back in time and choose again.

so while the answer is 100% correct, the question is academic.

most people who have not read a lot of textbooks have a difficult time thinking from a textbook point of view. most people who are only reading textbooks have a difficult time accepting why their thinking, although 100% correct, isn't applicable in the real world.

 

[IluvVol]

Quote from Lawrence Chan:

I am trying my best not to point fingers here because I think someone mixed up the usage of statistics ...

1. If there is one and only one event of choosing a cup and then an offer to switch, then it is all about luck and forget about probabilities.

i.e. Go tell the winner of a lottery ticket that he is an idiot wasting $ on something with such low probability.
He can tell you that every single $ he won is real.

2. If you are betting on a series of combo events stated in #1, then do the switch because it will pay off!

i.e. if you do not believe me, write a small program to conduct the test over, say, 1000 cases of random selections.

Trading is about consistencies in exploiting the edge you have. That is probability at work.

To add, yes it DOES pay off to switch even if you play the game only a single time. The probabilities of winning double from 1/3 to 2/3. Agreed, one single game does not guarantee that you win, you still lose with a probability of 1/3 after switching. But what do you prefer 1/3 or 2/3? ;-)

 

[Panurgo]

Quote from IluvVol:

Well you pointed your finger....at yourself. Please play this game with me. I even give you the honor to be the dealer. Your only tasks are a) to charge me 1 dollar for each game played, and pay me 2.9 if I pick the right cup with the coin (I give you 10 cents so you can earn a decent living) b) after I chose a cup you just pick one of the empty cups and take it out of the game and offer me to switch. I assure you that I will bankrupt you FOR SURE no matter what amount you start with, given we play sufficiently many games. You are the one who does not understand statistics thats for sure.

This is MY edge over you, and its very consistent, and yes, that is what trading is about, but you seem to be on the losing side of my trades LOL

Would you like to play with me this modification?
After the dealer has taken out the empy one I will pay an extra 2.9 if you don't switch and you pay me ONLY 1.5 if the coin is in the other cup. If you switch I will pay you nothing.

Want to play?

 

[bidask]

actually i did not read the question wrong. the question was written incorrectly.

"if you had a chance to change your decision..."

i outta this thread.

Quote from Batman28:

now let's suppose you choose one of the cups e.g. 2. He then asks given that you choose any of the cups (2 in our example) and he takes one of the remaining cups away (not the one you chose), and if you had a chance to change your decision, would you stick to the cup you initially selected (i.e. 2)?

[/B]

 

[jrlvnv]

Quote from IluvVol:

You really dont get it. Ok, here you go, because you obviously are too ignorant to read any of the links on the internet that explain this in DETAIL: YOu agree that when you pick one cup initially you have 1/3 prob of winning. You also agree that the other cups combined have a prob of 2/3 of winning, right? So if you now take one cup away from the 2 then the set of those previous two, now one, still has a prob of winning of 2/3. Does this ring any bell? Updated information changes probabilities. Are you really that irgnorant or you just try to piss off people? I simply dont get it. Why dont you simulate it in excel. Generate 1000 binomially distributed vars, "standardize" them so you end up with values of 1,2 or 3 symbolizing that the coin is under cup 1, 2, or 3. THen you again generate random numbers to decide which of the empty cups to take out of the game......the rest you can set up yourself. If you really cant or are too lazy to do then get lost. You waste others time.

Look I am not trying to piss anyone off but in my humble opinion it seems like you want your cake and to eat it also. Ok I feel the problem lies in your brain knowing too much information and fail to realize how simple the problem is. So I'll try and explain for the no so smart.

3 choices a,b,c all picked 100 times comes out the same. We pick one and then take away one. Ok now we have a 50/50 game. Hmmm what just happened. We have a new game now and you want to be payed 2.9 to 1 for a 50/50 game. Well I like those odds also but you fail to see were the problem lies. You run all the numbers you want. You will win 1/2 of the time whether you chose the same or not. Just because you switch from A to C you think something life changing will happen? If you chooce A and we take B out of the equation can we say that B had no part in the equation in the first place?

One last thing before beddy bye time....

If you always choose A and we always remove B do you seriously think you have a advantage on switching to C? If you can answer that question I'll be happy. Now lets take it a step further. Numbers 1-10. You pick a number.... i remove 8 wrong numbers just leaving the one you picked and one other one. You can switch to the other one if you like....... Do you expect to be paid 10-1 odds????? Hmmm don't think so. That would be like playing craps throwing one die and then choose what you want to bet on. First die is a 6.... well load up on the 12 since it pays 31-1..... Kinda of unfair asking someone to pay you 2.9 when its a even money payoff.

One last note. Look some people have a lot less experience in math then others. I post on things I like. I come up with a lot of math questions in a casino and it strikes my curiosity. This post striked my brain and I wanted to learn more about it. I did read the online links provided and saw the little cartoon. Just because I don't agree with it doesn't mean I come here and try and pick fights. Can't anyone have a different opinion even when one is wrong? Well nighty night everyone

 

[jumper]

Quote from Hydroblunt:

Uhm, Donald Trump's casino went Bankrupt.

isn't TRMP still trading? maybe, that's not the best example but the casino biz is booming. check out HET, MGM, LVS, WYNN, etc. the point was its all based on probability.

 

[faure]

The explanation I liked best was this: Take a 1000 cups with one coin. Choose one cup and the dealer (who knows where the coin is) removes 998 of the cups leaving us with 2 cups. Would it be logical to switch to the cup the dealer didn't take away or do you think you guessed right the first time and your cup has the coin? Hmmm...

 

[IluvVol]

Quote from Panurgo:

Would you like to play with me this modification?
After the dealer has taken out the empy one I will pay an extra 2.9 if you don't switch and you pay me ONLY 1.5 if the coin is in the other cup. If you switch I will pay you nothing.

Want to play?

Is the 1.5 a) the new price instead of 1.0 you charge for playing the game? Or b) is this in addition to the 1.0 I pay to play in case I dont switch and subsequently lose?

If a)
play but still switch

If b)
play but dont switch