General Poker Discussion Poker Forums

Check out this book: http://www.amazon.com/Drunkards-Walk-Randomness-Rules-Lives/dp/0375424040

You can get it used for $3-4. Another book that people often suggest is Fooled by Randomness, but I find that that one is a bit drier.

This makes sense to me. Others had the same point. Do you all agree with the notion that everything that has been in the past doesn't effect any future outcome?

Another thought would be that EV would always even out after XXXXXXXX number of hands.

EV doesn't always even out, you could theoretically run like shit for all eternity, but it is highly unlikely.

It's not a matter if we agree with probability or not, math is math.

See I don't understand this logic at all. Probably why I never pursued or should have pursued a career in math.

Which part? How do you think that it should work?

See I don't understand this logic at all. Probably why I never pursued or should have pursued a career in math.

All the long run means is that over a very large sample size, you will statistically run very close to your true EV--NOT that you will be "on top." You could actually be playing poorly and just been running hot for a few K hands.

After a few million hands, you don't suddenly start running at your EV...you will still have the same chance of variance in the short term. However a bad or good run over a few K hands (or a few hundred K) will not affect the total EV of a very large sample of a few million hands very much.

if you want to understand randomness and have fun doing it (for 1 hour), listen to this:

http://www.radiolab.org/2009/jun/15/

Which part? How do you think that it should work?

Well if things even out over x amount of time and you run bad for most of your time shouldn't you be able to expect to run good for your last bit of time? If we believe that things even out over time?

Well if things even out over x amount of time and you run bad for most of your time shouldn't you be able to expect to run good for your last bit of time? If we believe that things even out over time?

No, that assumes the cards have some kind of memory of your running bad previously. If you have a 60% chance to win an all-hand, and play the same hand a million times, the one-millionth-and-first time you play it, you will still have a 60% chance to win the hand. You could still hit that 40% and lose 4x in a row though.

The bigger your sample size, the more closely the entire sample will be towards the mathematical EV. A bad run in a million hand sample will not change the overall EV very much.

Well if things even out over x amount of time and you run bad for most of your time shouldn't you be able to expect to run good for your last bit of time? If we believe that things even out over time?

you're misinterpreting regression to the mean.

Thanks guys for the responses and the links they're all helping.

If you are a breakeven player and you set out to play 1M hands of 100NL, you're expectation is to end up with $0 in winnings at the end (imagine a world with no rake!). You might picture your future graph dipping up and down, but average out to $0 winnings in the end.

But....if you run particularly bad during the first 100k hands, say you are down -$3k, what is your future expectation from that point forward? It is no longer $0, it is -$3k. Imagine if you had decided to play 900k hands, not 1M, you would understand that your future winnings after those 900k would be $0, wouldn't you (see first paragraph)? The fact that you happened to have dumped -$3k after 100k hands, has no bearing on any future results ("the cards have no memory"), you are not expected to make or lose money from that point on, so your expectation is your winnings will remain at -$3k after those 900k hands.

PS No, I can't explain why a breakeven 100NL player would be planning to play 1M hands, it's just easier to calculate zero EV


The Moving Finger writes; and, having writ,
Moves on: nor all thy Piety nor Wit
Shall lure it back to cancel half a Line,
Nor all thy Tears wash out a Word of it.

Yeah I'll give my layman answer and hope I don't butcher it. From the mathematical perspective I believe you're square rooting (lol) the variance to get the standard deviation so as you get larger and larger numbers, then standard deviation will get smaller relative to the sample size.

This isn't to say things will even out or move closer to the mean in say a big blind sense. The more hands you play the further your results will generally get from the expectation, but the sample will increase at a faster rate so the relative distance from the mean (like... bb/100) will go down.

Eg. Take running 10BIs under EV over 20k hands and 200 BIs under EV over 2 million hands. The second one is a larger deviation from the expected value but the first is a much larger relative deviation based on the sample (5bb/100 vs 1bb/100). If you looked at a graph of the second one the EV and winning lines would be close together (assuming decent winner) but is actually much further away from the expected value. So bottom line is things only get closer in a relativistic sense (based on sample size) where as from an absolute sense they continue to get further and further apart, but at a diminishing pace... due to rooting (double lol) the variance. This is all covered way better in the tolerance series btw.

I'm not 100% sure though I suck at math so someone poke some holes in it.

All the world's a stage
And all the men and women merely players

http://www.youtube.com/watch?v=RjOqaD5tWB0

Just wanted to let you know that most of you are wrong.

I am now 20 Buy Ins below expected EB through 3 sessions. It has been insane!

Cards clearly have favorites among us and they do seem to remember your past luck even though you are tossing around coins to prove that it's not true.

I even think I heard the Jack of heart giggling as it brought another River disaster.

Just wanted to let you know that most of you are wrong.

I am now 20 Buy Ins below expected EB through 3 sessions. It has been insane!

Cards clearly have favorites among us and they do seem to remember your past luck even though you are tossing around coins to prove that it's not true.

I even think I heard the Jack of heart giggling as it brought another River disaster.

link to tommy angelo?

Am I doomed?

Yes. Yes you are. That's nothing to do with poker though.