The point is that nobody is going to bother to try to teach you the math in this thread.

I'm not asking to be taught any 'math.'

I'm asking a simple, one word, (or one number) answer, as to how many hands would be considered a worthwhile sample size. I'd have thought 10,000 all-in hands would be more than reasonable, but if someone knowledgeable such as yourself suggests differently then I'd be interested to hear your thoughts.

I think I'm going to check out Probability forum

You keep asking questions that any normal person would be able to look up themselves.

And you never do anything sensible even when answers are given.

People aren't going to keep playing up to your childish nonsense,

Well, it is hardly a secret!

It's there bold as brass in the list of forums.

How come you managed to find this one to whine about your poor results but couldn't find the one most apposite to your probability questions?


More likely you've just discovered a new forum to troll.

The probability will definitely be more polite, and I am genuinely curious to see how some of the stats freaks who do not visit this thread handle his routine for a while, but I am guessing they would just ignore him after a few of his weird vague posts, especially when they see he never makes an effort to save and produce any real HH data other than stuff he remembers/ makes up from the day.

Basically he would be a variation of the guys who ask about the martingale system in terms of the productivity of the thread.

I told him many times in the past he should post to the probability forum (and the psychology forum) so his claims of not knowing it exists is pretty par for his course.

Good luck probability guys!

As stated previously, I am not making up any results nor is my memory good enough to rely on to store numerous hand histories and percentages.

I have posted my question in the probability forum and an hour later there has been no response.

Perhaps someone wired ahead and told them you were coming.

Your data is cherry picked because you have chosen an arbitrary point within a long sequence of events to post in this thread about a cherry picked data point. I am bored of trying to explain this to you and others but I will resort to the coin tossing again.

The odds of tossing two heads in a row is 0.5 x 0.5 = 25% based on the way you are calculating it, but this is the odds of it happening in the next two tosses not the odds of it happening within a string of results. If you have done three tosses, the sequence could be (each of which is equally likely):

HHH *
HHT *
HTH
THH *
TTT
TTH
THT
HTT

So the odds of getting at least two heads in a row in this sequence of three tosses is 3 of 8 sequences (*) or 37.5%.

In 4 tosses:

HHHH (*)
HHHT (*)
HHTH (*)
HTHH (*)
THHH (*)
HHTT (*)
HTTH
TTHH (*)
THTH
HTHT
THHT (*)
TTTT
TTTH
TTHT
THTT
HTTT

So the odds of getting at least two heads in a row in this sequence of four tosses (I hope) is 8 of 16 sequences (*) or 50%.

So going back to your example of losing four 80/20 all ins in a row, the odds aren't simply a case of (1/5)^4 when in fact you have a huge string of 80/20 showdowns and you have selected the worst string that you could find (or the worst one you can remember). This happening when you started 6 max seems "a bit suspicious" to you but as I explained, if you had run well for a while you wouldnt have come here whining. Then the memorable sequence of bad results comes and you would be here whining "I ran good when I started the new game type to get me hooked and now I've been doomswicthed. Look!".

Worse still is that not only have you selected arbitrary times to post here (i.e. cherry picked strings of data) you are also selecting arbitrary types of data to post, in this case 80/20s but not mentioning 20/80s,60/40s or any other results that you may be running well in. So returning to the coin example, it is a situation similar to one where you have multiple coins, are flipping them for ages and then picking a particular sequence you want to complain about for a particular coin. The event which you are actually calling "odds of four heads in a row OMG (1/2)^4" is in fact "odds of an arbitrary worst run of results within a string of coin tosses over some set of multiple coins"

I don't know why I have wasted my time rambling on with elementary Maths to a Maths Graduate but you don't seem to understand what you have been told time and time again. Either

a) Choose what you are going to test for in advance, or
b) Analyse all the data you have

The kind of thinly veiled accusations you are making (basic stolen EV rather than elaborate equity balancing schemes) would probably be tested by somebody here if you post actual verified data, if you feel you aren't able to do it yourself.

Profitability in poker lays upon an edge of a knife in cash games.

One cold deck pre-loaded hand. An occasional software override to favor a big losing player to suck out that miracle card vs. a reg. That is all it takes to make a profitable player a loser. No one can overcome losing an extra buy in per 3000 hands. No one can overcome a pre-loaded cooler or software override per 3000 hands. How can someone prove it? All-in EV's for miracle suckouts. The other is far more complicated. You'd have to go through all your hands and see how many monster vs. monster hands are flopped. How many are in your favor and how many are against.

All you have to do is lose 1 out of 200 times more then your supposed to. If that 1 time is for your entire stack. Then it will be impossible to beat the rake (except high stakes depending on competition level)

Thank you for that shocking, breaking news.

Lock this thread up, Freak999 finally cracked the case.

I think you've misunderstood what I'm saying Bingo_Boy, in which case I can understand your comments about me lacking logic for someone with a maths degree etc.

I certainly wasn't picking a specific sequence of hands out of a large number of hand histories where I lost 4 hands in a row as big favourite and claiming that was really dodgy or (1/5)^4 in the 80/20 example you gave. That's not what I was doing, I understand that's not how probabilities work and that there would be nothing dodgy about that at all.

The results I've quoted are the ONLY results I've analyzed.

Several months ago, whilst playing heads up and whilst posting on here, I went away and started keeping track of my hand histories with the intention of analyzing these hands. However, as I was playing super turbos, (5 minute tournaments on average, playing around 30 of these an hour) and playing a large number of these in order to maximize my rakeback, I just didn't have the time to go back and analyze the hands at the end of the week or whatever, so I gave up on keeping track of the hand histories.

Last week, when I started playing the 6 man tournaments, at a rate of maybe 4 an hour, I felt it would be a more managable amount of hand histories to monitor and analyze, so I started doing so. So these are the only hand histories I have. I'm not cherry-picking these from a large sample as you seem to have thought.

So the odds of me losing my first four 80/20 situations in a row would indeed be (1/5)^4.

Also, I certainly wouldn't be coming here claiming I'd been 'doomswitched' if I ran badly at a later date, as you suggested. That's not a word I've ever used, it isn't even a word infact and you can't simply lump everyone who doesn't share your point of view in as 'riggies' and assume we all have the same opinions, that simply isn't the case.

Also, I'm not cherry-picking specific types of hand, I'm simply talking about the most common types of all-in situations that occur, as I explained in an earlier post and so far a lot of them, (perhaps all of them) are not running anywhere near the expected values.

As I said previously, I'll be happy to come on here and say what the results look like for a larger sample size, but this is all I've got at the moment.

That's something I was mentioning above, in terms of seeing a lot of 'carve-ups.' Hands where big hands, (which are going to be played for all a player's chips,) run into each other.

I'm not sure how we measure for how often this should happen and how often it is happening.

I understand there are lots of statistics that can be checked to see whether the dealing is fair, but there are also plenty more which are not so easy to check and which the sites could be using to pull a fast one.

Which ones and why are they not so easy to check?

You absolutely are, unless these are the only all-in hands you ever played in your life. You either have to look at your entire history, OR you have to decide at some point to run a test on your FUTURE hands and count nothing that happened prior to that decision. But what you cannot do is decide that these are the hands you want to look at (even as part of a go-forward series), AFTER you notice something unusual. That's kind of the definition of cherry picking.

There are a couple replies in the probability forum, the first is an interesting one that goes into nice detail about sample sizes. Not a bad read for everyone, though I assume most riggies and fake maths majors will ignore it, since really what can they say about it without sounding like a 3 year old in comparison.

http://forumserver.twoplustwo.com/25...e-size-918943/

Big hands running into each other is bad for the poker sites. So it'd be pretty funny if they were rigging the games to make that happen more often then it should.

Whoa now, thinking that the riggedness has to help the site, that's crazy talk.

Majority of grinder's all in EV is below expectation. ---IJS.

millions of hands and I average 1 buy below expectation on All-in EV for every 1000 hands. It's usually the dead money fishes or the extremely high BB/100 guys that are winning these all in's too much. The 2 players that kill my all-in EV graph. Lucky I had rakeback and other perks as well as being a prop. Otherwise there would have been no way I could possibly beat the rake.

It's only bad when it's two losers vs. two winners. If one of the losers is an ATM, poker addict, it's not bad for the site either. That player will just keep reloading = poker site win.

When it's a dead money fish vs. a winner and the dead fish wins, it's very good for the poker site.


There are definitely some sock puppets in this thread. Spin, spin, spin.

Yes. That's what I did. When I moved from heads-up to 6-man tournaments, I decided I was going to run a test on my future hands and started keeping track of my all-in hands.

Really ?

I'd love to hear your explanation for that, because I've got a fairly lenghty explanation for why the opposite of that is true.

If you are getting paid to post itt..please stand up!

It's about time we all realize that the people arguing in here everyday for the last how many years are indeed PAID posters. If you don't believe that then search google for "posters for hire". These people like spadebidder and monteroy etc etc are just shills getting paid some little amount to do someone elses dirty work.

I've met alot of people in my life and nobody is lonely enough to post in this thread all day everyday on a volunteer basis. 2+2 is in fact just a money making venture. They even hire people to do random hand analysis in the strategy forums.

Don't waste your time here.

1 buy-in / 1000 = 10bb/100 under EV over millions of hands. You must be the unluckiest person in the world!!!

That means you're currently several thousand buy-ins under expectation?!?! Assuming you play 0.50-1.00, you're running hundreds of thousands of dollars below what you should be winning?!?! The fact that you've still managed to beat the game means that you must be the second coming of Phil Ivey!!!!

I'd love to see your graph with AIEV. Better yet, post it in BBV, it'll be the sickest thing anyone has ever seen!!!!!!!

You do know that 2+2 has a subscribed threads page where you can see when there are new posts on threads you're interested in. Given how often this thread is bumped, its always at the top of my list. Do you understand now?