is it guaranteed that you will see your true expectation long term? is it impossible for chance to always go bad?

Lets say i keep flipping with a guy and i have a 60/40 edge , can i forever lose?

If its possible to always lose then why we think poker is a profitable game?

Also if the opposite is true , what proves that chance can never always go bad?

I'll take a stab. This has to do with your depth of understanding of probability and long run. And your understanding of the difference between probable/plausible and possible.

So, I'll ask you a question - Is it possible to throw a dime over a house intending for it to land in a standard beer bottle and it does land in that bottle? That answer is, I believe, yes. Is it plausible? That answer is, I believe, no.

The longer a run of bad luck is, the less likely it was to have happened from the start. And no matter how long you've been running bad, the current status of your luck is the same as it was just before you started your run bad.

The better you are at understanding this truth OR the deeper your comprehension of what was just said, the better equipped you are to deal with it's ramifications. Understanding will set you free - lack of understanding can cause a world of doom, gloom, and hurt.

This is a little like asking "if it's possible to fall down how do we know we can cross the street?" Yes, some people will fall on their way across the street and not making it across, even if they are fully capable of making it.

Some poker players will have luck that is worse than average and make less money than they should. Some will make more than they should. The longer it goes on, the less likely it is to have an extremely bad or good result.

But I bet there are plenty of people who went broke and quit who had the ability to win but ran bad.

You can, but the probability is 0 (not 0 after rounding, just zero period). Stated another way, there's literally an infinitesimal chance of it happening.

In a real lifetime he can absolutely lose "forever" i.e. never pull ahead cumulatively. I think that's what he meant. I don't think he meant in a theoretical infinity.

I was basically going to say what NewOldGuy said I guess. If the OP meant is it possible that he could lose forever, then it isn't infinity, but rather just however long he lives. Though even then, I would guess the chance is still extremely close to 0 (assuming an average lifespan - and I guess I was assuming an average life span for the country I live in, but I think just assuming an average lifespan for someone with regular internet access probably is about the same).

This seems like a kind of random walk problem. Weird things happen when you talk about, say, flipping a coin an infinite number of time. Like, you will eventually cross the zero mark into positive territory if you just play long enough.

Yes what newguy and lego said. In a finite lifetime there's a nonzero probability, but it's indescribably close to zero in the 60/40 scenario.
If it's possible for the player to always win in Roulette, why do we think Roulette is profitable for the house?

Kind of silly questions. In theory, I can play 10 lotto tickets every night for a week and hit the jackpot for millions every night for that week, but the probability is so remote that it's essentially impossible.

But the 60/40 question is nice one to show a sliding scale of chance. Let's say I have a 60% chance of winning an event. What are the chances of me losing more events than I win if I have a 60% chance of winning every time?

Well, for 1 event the chances of losing more is obviously 40%, and if we play 3 times, that's obviously 35.2%. Let's look at how the chances decrease the more we play by using binomial calculations. We'll use odd numbers so ties are impossible.

11 games - 24.7% (.2465)
21 games - 17.44% (.1744)
31 games - 12.84% (.1284)
41 games - 9.65% (.0965)
51 games- 7.35% (.0735)

101 games - 2.09% (.0209)
151 games - 0.64% (.0064)
201 games- 0.21% (.0021) (1 in 476)
501 games- 0.0003% (.000003) (1 in 333,333)

Obviously, these illustrate the chances of the probabilities for a set amount of games before we begin. The true odds will change after every game. For example, I might only have a 12.84% chance of losing in a 31 game block, but if I was to start that block and lose the first 5 games, my chances of losing that block raises to 47.88%.

Bottom line, enough volume will almost always balance variance, but variance can be a bumpy ride. Someone in the casino business years ago made the smart decision to post what numbers and their color has hit in the last 15 spins of a roulette wheel. Why? Because suckers who don't understand variance would see 12 out of the last 15 spins were red, so they bet more money than they should thinking Black has to start coming up more because it's a 50-50 chance (minus the 0 of course). So when Red comes up another 12 out of 15, they get buried and take out a line of credit to bet Black even larger, only to lose their house because they couldn't grasp how variance works.

I recently hit back to back quads on a video poker game called pick'em poker. The chances of hitting quads on any spin (playing proper strategy) is 2,360-1. The chances of hitting back to back quads is around 5.5 million to 1.

It was the only time that ever happened to me, and most likely I'll never see that happen again. However, considering i have played over 1.6 million hands of that game, it's really not that amazing when taken into the context of the bigger picture.

If someone were to flip heads 15 times in a row on a coin, that's a remarkable event that has a 32,768-1 chance of occurring. However, if I was flipping coins at the rate of 50 flips a minute for 18 hours a day, there is a 56.1% chance I'll have at least 1 streak of 15 heads in a row that day.

So how does all this relate to poker? Simple. If you play a lot of poker, you'll experience hot streaks and bad runs that will boggle the mind, but if you don't lose your mind, and can keep playing well enough to maintain an edge in the games you play, you'll do fine if you play enough games,hands, tournaments, or whatever your main game variant is. Just make sure you're prepared for long stretches of bad luck and losing so that when it happens, you don't go insane and/or broke.

That is a great way to state it. I was going to talk about sets of measure 0, but you are much better at simplifying things than I am.

As an analogy, the probability a clock will stop at any specific time,say exactly at 10 o'clock is 0 yet obviously it will stop at some time.

Lately yes

In fact you will do this an infinite number of times (unless the coin is weighted).

The chances of a clock stopping at exactly 10 o'clock is not zero. It's .0014, and at only 719-1 odds, hardly an earth shattering occurrence.

If we assume 1-second resolution, the clock face has 43,200 places it can stop. 720 would be 1-minute resolution. But if we assumed continuous resolution, then the chance actually is zero.

The chance is never zero. You can add all the resolutions you want. If you have any instrument measuring time in a looping 24 hour cycle, and that clock is powered by some method which will eventually cease working, then there is always a chance, no matter how remote, that it will stop working while it reads exactly 10'o clock or 10:00, or 10:00.00.

No the probability of landing on an exact number of a continuous distribution is 0. This is generally taught in introductory probability and stats. You have to remember the all probabilities (of continuous distributions) are integrals. The integral between t and t is always 0. I have taught this a few times, so I can give you some good reference material if you want to read about the different between discrete and continuous distribution.

Sure, it's possible for it to land exactly on 10 o'clock, it just happens with a probability of 0. You could also say it happens almost never which is really just a clever restatement of the situation.

Think about it this way - if the clock measures the time continuously, ie to an infinite decimal place, then what are the odds of seeing any specific time? 1 divided by infinity. Which equals exactly 0.

Probability zero does not mean the same thing as impossible. Yes the probability is 0 to land on any exact value of a continuous distribution. But it does in fact have to land somewhere.

This of course is correct.

It's already been explained why we still define that probability as zero, which doesn't mean it's impossible. But I thought it might be helpful to note that in real life, of course physical infinities do not exist and are merely a thought experiment. And it's very doubtful that continuous distributions can exist in the physical world either, it's just a matter of the degree of resolution that is possible.

So yes, any place on the clock face that we can actually define or measure in the real world, has a non-zero probability, because our ability to resolve the placement is limited, and even if it weren't limited, there is a discreteness to it at some level eventually.

I think it's more likely to stop at 10:08 and 42secs.

Don't ask me to prove it scientifically. Just a gut feeling. Might have something to do with that Seiko thing.

For some reason I enjoy thinking about these types of things - they're just thought provoking in my opinion.

I 100% agree with this. But it's weird when you include the reflexive property of equality.

P(impossible event) = 0

But P(event) = 0 does not imply the event is impossible.

It's weird that the probability of an impossible event is 0 but a probability of 0 does not mean the even is impossible.

This comes to life in this discussion.

Let x be a random variable which denotes an amount of water measured in volume, say ounces. x is a continuous random variable.

Let x be a random variable which denotes an amount of water measured in molecules. x is a discrete random variable.

Is "amount of water" discrete or continuous?

In my opinion, it's both OR it depends on how you choose to think about it. OR may be it depends on how you choose to define "amount" (volume or individual molecules).

Take any exact amount of water. Now add an amount that is less than the volume of one atom. Or one quark. Or something smaller not yet known. Unless you can do so, the volume is discrete.

The same number of 'molecules' can have more than one volume.

http://www.ehow.com/how-does_5185456...t-heated_.html

Under the guise of splitting hairs - I'd say volume is continuous. Unless volume takes quantum like leaps as electrons do.

But I see what you're saying - thanks!